I found as I was preparing to teach "Introduction to thermodynamics" that there are varying definitions or examples of pure substances. For instance the text book I have to use "Thermodynamics an engineering approach, Cengel and Boles" defines a pure substance as one that has the same chemical composition throughout. Air in the gaseous form is cited as an example. Air in its liquefied form is NOT since it seperates into it's various components.

The Dummies book on Chemistry suggests that sugar can be considered a pure substance as it has the same chemical composition throughout. And water can be considered a pure substance for the same reason.

So would a homogeneous mixture of sugar and water be considered a pure substance? After all, a homogeneous mixture has "identical properties" throughout it's phases.

Yet another definition of a pure substance suggests that any material made of the same kind of atoms is a pure substance. So individual elements in the periodic table and stuff made "purely" of them are pure substances. Other stuff is just homogeneous or heterogeneous mixtures.

So are pure substances (irrespective of whether we view them as elements, compounds or mixtures) just very difficult to break down into their individual components?

Or would the definition of a pure substance change depending on who you speak to (chemists, phycisists, thermodynamicists)?

  • 2
    $\begingroup$ Air is not a pure substance; it is a mixture. $\endgroup$ – Nick Jan 4 '13 at 12:17
  • $\begingroup$ @Nick but the constituents of air cannot be extracted without a lot of effort. So in case a lot of energy is required to extract constituents, can it be described as a pure substance? $\endgroup$ – dearN Jan 4 '13 at 19:42
  • 2
    $\begingroup$ no, a pure substance is not a mixture. If you can separate substances from a mixture using a centrifuge, chromatography, or any other physical method then they are not pure. $\endgroup$ – Nick Jan 4 '13 at 19:48
  • $\begingroup$ I am unable to choose a right answer to this rather subjective question. Any advice? $\endgroup$ – dearN Jan 7 '13 at 23:18
  • 2
    $\begingroup$ @BenNorris 's answer seems correct to me. Engineers might use a different definition if it suits their purposes, but Chemistry defines a pure substance precisely. $\endgroup$ – Nick Jan 8 '13 at 9:13

"Thermodynamics an engineering approach, Cengel and Boles" defines a pure substance as one that has the same chemical composition throughout.

This is the correct definition of a pure substance. However, air, regardless of phase, is not a pure substance.

All matter is categorized as either a "pure" substance or a mixture. The word "pure" in front of "substance" is unnecessary, since the definition of substance implies purity. A substance is a sample that is chemically uniform in composition. In other words, all particles of that sample are the same chemically (whether they are atoms or molecules or ionic lattices or what have you). Substances cannot be separated into simpler components by any physical process (i.e. without breaking chemical bonds). Mixtures are samples that contain two or more substances. All mixtures can be physically separated, although the means to do so may be neither easy nor practical.

Air is a mixture of nitrogen ($\ce{N2}$), oxygen ($\ce{O2}$), carbon dioxide ($\ce{CO2}$), argon ($\ce{Ar}$), water vapor, and traces of other gases. Each of these gases separately is a substance. Air can be easily (but not in the comfort of your own home) separated into its constituent parts. Cooling air will cause the constituent components to condense out of it at progressively lower temperature: first water vapor, and then $\ce{CO2}$ (195 K), followed by $\ce{Kr}$ (120 K), $\ce{O2}$ and $\ce{CH4}$ (90 K), $\ce{Ar}$ (87 K), $\ce{N2}$, $\ce{Ne}$ (27 K), $\ce{H2}$ (20 K), and $\ce{He}$ (4 K).

As each substance condenses out of air it can be physically separated from the remaining gas, reducing the complexity of the mixture. Each substance, if returned to the gaseous state separate from the original mixture is physically and chemically distinct from the original mixture (or whatever is left of it). Such physical separations do not need to be easy to make air a mixture, they need to be possible.

A pure substance does not exhibit this behavior. When pure water is cooled, it undergoes a phase change to ice, which is physically separable from liquid water. However, the solid and liquid samples are still water. You have not reduced the complexity of the sample, just changed its form. If you allow the separated solid water to melt, it will be physically and chemically indistinguishable from the original water sample. Pure water cannot be separated into other chemically distinct substances without performing a chemical reaction, which means breaking and forming bonds between atoms.

The majority of matter that we interact with in our everyday lives is comprised of mixtures.

We encounter a few pure substances in the kitchen. Table sugar is a single compound (sucrose). So is table salt (sodium chloride), baking soda (sodium bicarbonate), and cream of tartar (potassium bitartrate). Water is a pure substance, so long as it has been purified. Most water that we encounter has other substances dissolved in it.

Gold (as long as it is 24K) is a pure elemental substance. So is diamond (carbon), graphite (also carbon), and the neon in neon lights. All minerals are pure substances, including asbestos. Glass is a pure substance. Most precious stones are pure substances.

  • $\begingroup$ Van Wylen in his textbook "Fundamentals of classical thermodynamics" defines air as a pure substance as long as it is in it's gaseous form. When liquified, it seperates into it's components. $\endgroup$ – dearN Jan 5 '13 at 13:47
  • $\begingroup$ Air is a homogeneous mixture of several gaseous substances. If air is a pure substance, so too is diet cola (also a homogeneous mixture). $\endgroup$ – Ben Norris Jan 5 '13 at 13:53
  • $\begingroup$ Yes, the punch line of Wan Vylen's definition is that air CANNOT BE SEPARATED into it's constituents easily in the gaseous form. I Think just opening a can of diet cola/regular cola starts releasing CO2 bubbles... $\endgroup$ – dearN Jan 5 '13 at 13:57
  • 1
    $\begingroup$ I guess the definition is quite subjective. $\endgroup$ – dearN Jan 5 '13 at 14:33
  • 1
    $\begingroup$ Only the definition of "all particles being of the same chemical identity" is no good, because that would render all ionic compounds mixtures. Even purified water is strictly a mixture of $\ce{H_2O}$ but also $\ce{H_3O^+}$ and $\ce{OH^-}$ (even you describe those, quantum mechanically, in terms of excitation particles). You don't get around the definition of being not easily physically separable, which is indeed subjective – but so are many definitions in chemistry. $\endgroup$ – leftaroundabout Jan 6 '13 at 12:32

Simply, a pure substance is a substance which is composed by molecules or directly by atoms of only one kind. Diamond is a pure substance because it is directly made up of carbon atoms. Salt water is not because it contains sodium chloride and water.

But there is almost no 100% pure substance in this world (Exceptions exist, as there can be only one or two molecule and you can call that a substance), thus we normally recognize relatively pure substances as pure substances.

  • $\begingroup$ The definition of a "pure substance" seems to be quite subjective. I have found a couple of these varying instances now. One is like I have noted in Cengel and Boles. Van Wylen in his text book says air, for instance, can be treated as a pure substance as it cannot be separated into it's constituents in the gaseous form. In liquified form, air is NOT a pure substance. $\endgroup$ – dearN Jan 5 '13 at 13:49

It is quite possible for a material consisting entirely of one element to be "impure" - for example, a mixture of carbon nanotubes, graphite and buckyballs could hardly be called "pure" despite being made of carbon only. For some applications, even a mixture of different types or lengths of carbon nanotube isn't "pure" enough.

On the other hand, you can buy "pure" orange juice in the shops that is neither entirely one element nor entirely one molecule. In fact, "pure (from concentrate)" orange juice isn't even entirely made up of stuff that was squeezed out of an orange. So "purity" is in effect a matter of how you define it.

For thermodynamics purposes, "purity" is a kind of order - if this order is lost by mixing two pure substances together, it takes work to un-mix them. Conversely, with a carefully designed machine it is possible to harness the entropic "force" in favour of mixing to generate useful work. ATPase, the enzyme which performs the final stage of converting food energy into ATP, the internal energy "currency" used by chemical processes in cells, is the most widespread example (although in this case it was evolved rather than designed). As a crude summary, it uses the energy released by mixing acid and water to drive a rotor, which as it turns forces ADP and phosphate together making ATP. (The acid is generated in earlier stages of the process.)


In the lab, we refer to a pure substance as any substance with no impurities.

That is to say: pure water is 100 % $\ce{H2O}$ with nothing else in it; pure acetone is 100 % acetone.

We do however frequently use the term with respect to known mixtures. One can say we have pure clay despite the clay containing several KNOWN components because there are (practically) no impurities that have not been identified or would upset our experiments.

I would be careful describing a mixture as pure, however, unless you know everyone is on the same page. Pure tap water sounds absurd but pure air, oil, product, or rubber doesn’t.


Basically, a pure substance in this context means a substance that is always available. As in the expression of rate law, if a solvent is always available,say water. Then rate won't depend on the concentration of water , as it is always available. Hence its activity can be considered as 1. Certain textbooks refer to pure substances in physical chemistry in this sense.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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