A fundamental difference between a metal and a semiconductor is that the former is unipolar, whereas a semiconductor is bipolar.

[Section 2.5, Millmans Integrated electronics]

Can somebody explain why is it considered so, as after all electrons are responsible for the current which leave the one end(-) of voltage source and enter the another end(+) in same quantity. I am unable to figure out holes shifting into the -ve terminal and thus act as bipolar current.

  • \$\begingroup\$ This is a rather vague statement. More context is required to determine what the author wants to say. \$\endgroup\$
    – mng
    Jun 10, 2012 at 19:17

4 Answers 4


That's not how I learned it. It has to do with energy gaps between valence band where the electrons at the highest energy are, and the conduction band, where an electron can come free of its atom. In metals these bands overlap and electrons move free within the metal's lattice, and that's what gives metals their typical shine.

Pure semiconductors are isolators at cryo-temperatures. But doping them with N-type material will cause doping atoms to bond with the semiconductor, and then there's one superfluous electron which, like in metals, can move freely through the lattice, and thus conduct electricity.

A useful way to visualize the difference between conductors, insulators and semiconductors is to plot the available energies for electrons in the materials. Instead of having discrete energies as in the case of free atoms, the available energy states form bands. Crucial to the conduction process is whether or not there are electrons in the conduction band. In insulators the electrons in the valence band are separated by a large gap from the conduction band, in conductors like metals the valence band overlaps the conduction band, and in semiconductors there is a small enough gap between the valence and conduction bands that thermal or other excitations can bridge the gap. With such a small gap, the presence of a small percentage of a doping material can increase conductivity dramatically.

An important parameter in the band theory is the Fermi level, the top of the available electron energy levels at low temperatures. The position of the Fermi level with the relation to the conduction band is a crucial factor in determining electrical properties.

enter image description here

(from this excellent site)

  • \$\begingroup\$ Semiconductors are isolators at very low temperatures, and conduct better with rising temperatures. \$\endgroup\$
    – posipiet
    Jun 10, 2012 at 17:15
  • \$\begingroup\$ @posipiet - right, I was still in the cryo-lab :-) \$\endgroup\$
    – stevenvh
    Jun 10, 2012 at 17:18
  • \$\begingroup\$ How is the overlapping of conduction and the valence bands related to the shininess of the metals? Shouldn't it be related to the photons the material can emit/reflect? (Sorry, I realize this is a really old thread) \$\endgroup\$ Nov 8, 2014 at 7:14

I see at least two correct answers here, however I get a feeling (based on the way the question was phrased) that it might be too complicated to understand this in terms of "valence/conduction bands overlap/gap".

The sentence you cited should be considered only in the whole context of the text. It does not make much sense standalone.

Let me try to describe the difference in simpler words.

The major difference between conductors and insulators is the amount of "free" conduction electrons - these electrons can be affected by the externally applied electric field, thus contributing to current flow.

enter image description here

  • Conductors have many conduction electrons, therefore even low electric field (low voltage) can cause a large current. The other way to say the same thing is that the conductivity of conductors is high.
  • Insulators have few conduction electrons, therefore their conductivity is low.

The definitions of "low" conductivity and "high" conductivity is heuristic, and it is possible that a material, which is a good insulator in one application, will be considered a conductor in some other application.

Most metals are very good conductors.

However, even an insulating material can be a very good conductor. Take an air for example - it is a very good insulator, but can become a very good conductor when the molecules are ionized. This ionization process and the subsequent conduction can be seen even by naked eye in forms of lightnings and sparks.

It usually takes a lot of energy to turn an insulator into conductor. This energy can be obtained from a strong electric field, high temperatures, etc... The exact amount of energy required is material specific (this energy is sometimes referred to as Band Gap energy). This brings us to semiconductors.

enter image description here

Semiconductors are insulators which require relatively low energies in order to "kick" the electrons to a conduction state. When electrons become "free", the conductivity of semiconductor rises. For example: semiconductor thermometer having an accuracy of 1\$^{\circ}\$C will change its conductivity in a measurable manner for temperature differences as low as 1\$^{\circ}\$C.

Energy, energy, energy... How is it related to the distinction between conductors and insulators? This energy, when transferred to "non-free" electrons, causes them to become a conduction electrons. This energy is used to "kick the electrons out of their permanent bonds" - once the electron is "kicked out", it becomes "free" and can be affected by electric field and contribute to current flow.

The above describes how do conduction electrons emerge in semiconductors, but there is one more mechanism which contributes to conductivity: when some electron is "kicked out" and becomes a conduction electron, it leaves an empty space behind (one empty electronic state in the vicinity of atom's nuclei). This state can be accommodated by another "non-free" electron "jumping" from its current position to this empty state. Why would it "jump"? Most importantly: due to external electric field which causes the electron to "want" to "jump".

The above "jump" populates the empty state, but creates another empty state in other place. You can see this empty state as "moving by himself". Since the motion of this state is in opposite direction compared to the "jumps" of the electrons, we can see it as being affected by the same electric field, but having a positive electric charge.

enter image description here

The above empty state which can "move around" under influence of the electric field and having equivalent positive charge is called a hole. Both negative conduction electrons and positive holes can contribute to current flow in semiconductors, therefore the latter are called bi-polar materials.

You can think of metals as having all the conduction electrons always present - no need to provide any additional energy in order to "kick" them to "freedom". Since no additional electrons can be produced in metals - no holes left behind. Therefore, in metals only electrons contribute to current flow, and you can say that metals are uni-polar materials.

Note that in both cases the only charge carriers are electrons. However, it is much more convenient to think of a positive hole moving towards (-) terminal, than thinking of many electrons "jumping" one after another towards (+) terminal. This should not be too hard for any electrical engineer since all of us are accustomed to treat the current as a motion of positive charges, neglecting the fact that these are electrons that in motion.

Hope this helps.


The explanation in the book is incomplete: The major difference between metals and semiconductors is how the valence band is populated. A nice representation is here wikipedia.

In semiconductors, the valence band and conduction band do not overlap, allowing holes to exist in the valence band as soon as electrons jump into the conduction band. Since holes represent a positive charge, they move in the opposite direction and "get filled" with electrons at the cathode.

In metals, the valence band overlaps the conduction band. This makes the valence band, in which holes would appear if electrons jumped into the conduction band, fully populated at all times. Therefore, only electrons move inside metals.

  • \$\begingroup\$ I couldn't get the argument of why holes could not exist in the valence band of the metals? \$\endgroup\$ Nov 8, 2014 at 7:18
  • \$\begingroup\$ Electrons move in the insulative materials also, just much much less. \$\endgroup\$
    – Voltage Spike
    May 17, 2021 at 20:53

Section 2.5, Millmans Integrated electronics]

A fundamental difference between a metal and a semiconductor is that the former is unipolar [conducts current by means of charges (electrons) of one sign only], where as a semiconductor is bipolar (contains two charge-carrying "particles" of opposite sign).

The above paragraph is copy-pasted from the refered text book to clarify the question as it is originally written it misses some important words.

It is understood in the context that unipolar means that conduction of electric current is only due to one type of carriers and bipolar means that conduction of electric current is due to two types of carriers.

It is well known that holes can be imagined as electrons moving in opposite direction with respect to direction of holes but still the behaviour of a hole is distinct from that of electron as suggested by quantum theory which is well proved by Hall's experiment. This very experiment is sufficient to understand that electric current through a semiconductor(not through the voltage source) is carried by both electrons and holes.

I am unable to figure out holes shifting into the -ve terminal and thus act as bipolar current.

In semiconductor diodes there are very few holes in the n-doped region of semiconductor generated due to electron-hole pair recombination.

(let us assume diode is forward biased for the time being)

enter image description here
(source: gsu.edu)

Here forward biased battery is attached in such a way that the negative terminal is connected to the n-side of diode and positive terminal to the p-side of diode. Now you misunderstood that only electrons from metal are given into the n-side which are caught by the positive terminal of the battery as electrons flow through the p-side. In fact holes contained in the p-side also flow through the n-side and are captured by the negative terminal of the battery.

Hence, for a forward-biased diode, the holes cross the junction from the p-type into the n-type region, where they constitute an injected minority current.


The holes in the n-side = a few thermally generated holes in the n-region + the holes came from the p-side of diode

And these holes will diffuse from the p-side to the negative terminal.

Usually we use copper wires to connect a diode with battery and in such a situation there is always a Metal-Semiconductor junction of very little resistance and for sure these holes will diffuse across the M-S junction and will eventually combine with the electrons in Metal due very high recombination chances.

  • \$\begingroup\$ What you are describing are doped semiconductors forming a PN junction. As I understand the question OP asks about pure semiconductors. \$\endgroup\$
    – amadeus
    Dec 3, 2013 at 10:21
  • 1
    \$\begingroup\$ anupam, I don't see any particular adversion to your answers. I really appreciate your enthusiasm and motivation, and I think that you can be a very good contributor for the site. I'd just like to give you a tip: don't rush into posting an answer before being sure that you make a clear point in it. Take your time to check spelling, formatting, consistency and avoid excess of abbreviations; here a clear language is much more appreciated than chat-like abbreviations. Also, expect some criticism (there's a lot of knowledgeable people here) and be open about looking again at your posts. \$\endgroup\$
    – clabacchio
    Dec 3, 2013 at 13:07

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.