I know in chemistry metals are a class of elements on the periodic table, but in physics metal is more like a state of matter. All of the elements that are called metals on the periodic table are metals under normal conditions.

Apart from the 4 classical states of matter, Wikipedia lists a lot more, not only extreme ones, like quark-gluon plasma, BEC, superfluids, supersolids, but also liquid crystals, states that just have different magnetic properties and even glass is considered as a unique state. Metal isn't on the list, nor is it included as a state of matter in any other literature I've managed to find.

The reason I'm confused about this is because elements that are nonmetallic actually have metallic states at high pressures, including hydrogen, helium, oxygen, carbon, other heavier nonmetals. Even water, a compound, has a metallic state at 4000K, 100GPa. Metals have nonmetallic allotropes too.

Also, the primary difference between a metal and a nonmetal is analogous to the difference between a gas and plasma, which are different states of matter.

Another state of matter was discovered not long ago called the Jahn-Teller Metal. If this is classified as a state of matter, how come metal can't be classified as one too?

  • $\begingroup$ All solids have a band structure. Some are metals, some semi-metals, some semiconductors, and some insulators. But, they are all solid. Further, note that many liquids are metallic as well, so being metal is not a thermodynamically distinguishing feature. $\endgroup$
    – Jon Custer
    Commented Jan 25, 2018 at 13:53

3 Answers 3


The problem is just that there are two different meanings of "state of matter"- the commonplace meaning, and the meaning used by physicists.

The commonplace meaning is that a state of matter is either solid, liquid, or gas. For some reason, plasma is now also often included. The definitions given usually involve rigidity and whether the substance fills its container.

The physicist meaning of state of matter is more like any substance with a given set of properties*, and so it has many more possibilities. For example, certain solids can change between being magnetic and being non-magnetic at a critical temperature, called the Curie temperature. Even though both are solids, they are still considered different states of matter in the physicist sense. Furthermore, in some cases the transition from liquid to gas is really a smooth crossover without any sharp transition, and in this case the two aren't really considered to be separate phases in the physicist sense. So the definitions are really quite different- one isn't just a generalization of the other. Unfortunately, popular descriptions almost always mix the two up, especially in a list like the Wikipedia list of phases.

So, then, is a metal a phase of matter? In the commonplace sense, it is not- the possible phases are solid, liquid, and gas, and metals can be either solid or liquid. However, in the physicist sense metals are indeed a phase, or depending on the system there may be multiple metallic phases. The study of metal-insulator transitions of various kinds is a major focus of condensed matter physics. The defining characteristic of a metal has varied somewhat over time and depending on context, but it is often either (1) A decreasing electric resistivity with decreasing temperature, (2) Extended (as opposed to localized) single-electron wavefunctions, or (3) A gapless (as opposed to gapped) energy spectrum.

*the formal definition generally involves some parameter that exists only within one of the phases, called an "order parameter"


It all really depends on how you define a state of matter, and there is no clear definition at all. For example, if you look on Wikipedia you will find this "definition"

In physics, a state of matter is one of the distinct forms in which matter can exist.

which gives almost no information (what exactly is meant by "form" here, and how do we distinguish them?).

However (as also reported in the Wikipedia article), there is another, more precise way to define a state of matter, which is the one that I personally prefer.

Indeed, while defining a state of matter can be complicated, defining a phase transition is relatively easier, since phase transition are characterized by well-defined particular behaviors (divergences, discontinuities, power-law behaviors with specific critical exponents...) of certain thermodynamic quantities (1). With this definition, we can say that some material went from a state of matter to another when it underwent a phase transition.

If we adopt this definition, then in some cases we can consider the "metallic state" as a state of matter. A recent work (2) has for example pointed out that the insulator-metal transition in hydrogen has the characteristics of a first order phase transition.

Therefore, it really seems like in some cases you could think of a metal as a distinct state of matter.

(1) I am including both first and second order phase transition in this very generic definition.

(2) Evidence of a first-order phase transition to metallic hydrogen, Mohamed Zaghoo, Ashkan Salamat, and Isaac F. Silvera Phys. Rev. B 93, 155128 (2016).

  • $\begingroup$ I like this answer. There is also the transition of solid oxygen to metallic at 96GPa journals.aps.org/prl/abstract/10.1103/PhysRevLett.74.4690 but I don't know if this is a first order transition or not. $\endgroup$
    – Curiosity
    Commented Jan 25, 2018 at 15:29
  • $\begingroup$ Using phase transitions as a defining factor (which has its own problems) means that plasma is not a distinct state of matter as there is no phase transition from gas to plasma. On the other hand, there is a known second order phase transition in some metals at the Curie temperature in addition to the long-suspected first order phase transition between non-metallic to metallic hydrogen. $\endgroup$ Commented Jan 25, 2018 at 15:58
  • $\begingroup$ @DavidHammen Don't know much about plasmas, but in this answer it is suggested that the gas to plasma transition could be a 2nd order phase transition. Do you have any reference where it is stated that it is definitely not a phase transition? I'm asking out of curiosity because I really don't know the answer. $\endgroup$
    – valerio
    Commented Jan 25, 2018 at 16:28
  • 1
    $\begingroup$ @valerio92 - Typically a second order phase transition means the transition involves a non-continuous second derivative. Plasmas do not exhibit that kind of behavior. Plasma physicists are wont to inflate the importance of their field. Their field is important, as is. It does not need a phony designation as a fourth phase of matter. $\endgroup$ Commented Jan 25, 2018 at 16:33
  • $\begingroup$ @DavidHammen Well, I agree with you there. If there is no reason to call it a separate state of matter, then let's just not do it. Regarding the definition of state of matter based on phase transitions, I am aware that it has its own problems, but at least we can base it on some quantitative consideration. $\endgroup$
    – valerio
    Commented Jan 25, 2018 at 16:36

My thinking is that states of matter are characterized by the behavior of the atoms or other basic constituents of a substance. In contrast, whether the substance is metallic (or insulating, or semiconducting, or superconducting) is an electronic property, classifying the behavior of the electrons. Naturally, the electronic properties are highly dependent on the configuration and behavior of the atoms. That's why, e.g., some allotropes of carbon are metallic (metallic nanotubes), while others are insulating (diamond), and still others are zero-bandgap semiconductors (various graphene thicknesses). Their atoms are arranged in different ways (and exist in differing numbers of spatial dimensions: 1, 2, 3!), leading to different electronic properties. But they are all solids because their atoms remain fixed in place as they do in solids.

If you look at your list of different states of matter, only the magnetic states can be considered a description of the electronic properties of a substance rather than of the atoms--the exception that proves the rule! For after all, electrons themselves can be modeled as existing as one of the states of matter (ever heard of Fermi liquid theory or the hydrodynamic theories of conduction?). Metals aren't another form of matter, the electrons in them are just funny liquids!

Anyway, don't get too caught up in these sorts of classifications. There's no governing body of classifying states of matter. The dirty little secret of this sort of thing is that we're all just making it up as we go along!

  • $\begingroup$ Good point, but then what about plasma? It's considered one of the classic states, but the only difference between it and a gas is the electrons being freed up. $\endgroup$
    – Curiosity
    Commented Jan 25, 2018 at 15:11
  • $\begingroup$ One reason I think this is interesting is the possible implications for chemistry if metallicity is treated as a variable instead of an element type. $\endgroup$
    – Curiosity
    Commented Jan 25, 2018 at 15:32
  • 2
    $\begingroup$ @Curiosity - I, for one, do not consider plasma to be a distinct state of matter. Plasmas are gases with some ill-specified fraction of the components being ionized. Any non-ideal gas at a temperature above absolute zero will have some portion of its constituents being ions. This fraction rises smoothly as the temperature increases. There is no phase transition, of any order, from gas to plasma. $\endgroup$ Commented Jan 25, 2018 at 16:06
  • $\begingroup$ I always thought this too, but it seems to be a well established thing now. So my thinking is why not both, or neither. $\endgroup$
    – Curiosity
    Commented Jan 25, 2018 at 16:37
  • 1
    $\begingroup$ @Curiosity See, my point is that this classification parsing ends up being silly semantics. But I stand by my (unclear?) assertion that you need to ask "what are the basic constituents that I'm talking about?" If you're considering the mass substance of the material, then you're talking about atoms, and the electronic properties are secondary. As I said, you could explicitly consider electrons, which have their own "states of matter". Plasmas are consistent with my assertion since we're talking about atoms, the atoms themselves are ionized, and they behave differently than a simple gas. $\endgroup$
    – Gilbert
    Commented Jan 25, 2018 at 17:15

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