Separation axiom

 For the axiom of set theory, see axiom schema of separation.
In topology and related fields of mathematics, there are several restrictions that one often makes on the kinds of topological spaces that one wishes to consider. Some of these restrictions are given by the separation axioms. These are sometimes called Tychonoff separation axioms after Andrey Tychonoff.
The separation axioms are axioms only in the sense that, when defining the notion of topological space, you could add these conditions as extra axioms to get a more restricted notion of what a topological space is. The modern approach is to fix once and for all the axiomatization of topological space and then speak of kinds of topological spaces. However, the term "separation axiom" has stuck. The separation axioms are denoted with the letter "T" after the German "Trennung", which means separation.
The precise meanings of the terms associated with the separation axioms has varied over time, as explained in History of the separation axioms. Especially when reading older literature, be sure to get the authors' definition of each condition mentioned to make sure that you know exactly what they mean.
Before we define the spaces described by the separation axioms, we need to define some terminology in order to give concrete meaning to the concept of separation.
Contents 
Separated sets and topologically distinguishable points
The separation axioms are about the use of topological means to distinguish disjoint sets and distinct points. It's not enough for elements of a topological space to be distinct; we may want them to be topologically distinguishable. Similarly, it's not enough for subsets of a topological space to be disjoint; we may want them to be separated (in any of various ways). The separation axioms all say, in one way or another, that points or sets that are distinguishable or separated in some weak sense must also be separated in some stronger sense.
Let X be a topological space. Then two subsets A and B of X are separated if each is disjoint from the other's closure. Note that any two separated sets must be disjoint.
There are several stronger forms of separatedness for sets; they are in order: separated by neighbourhoods; separated by closed neighbourhoods, separated by a function, and separated precisely by a function. These are defined and discussed in the article Separated sets.
Don't confuse separated sets with separated spaces, which are defined below.
Note that we sometimes use the terminology of separated sets to refer to points; in that situation, we're really talking about the singleton set {x} rather than the point x. If A and B are open and disjoint, then they must be separated by neighbourhoods; just take U := A and V := B. For this reason, many separation axioms refer specifically to closed sets.
Two points x and y in X are topologically distinguishable if they don't have exactly the same neighbourhoods. Note that if two points are topologically distinguishable, then certainly they are distinct. Furthermore, if the x and y are separated (that is if the singletons {x} and {y} are separated), then they are also topologically distinguishable. For more on topologically distinguishable points, see Topological distinguishability.
Definitions of the axioms
Many of these names have alternative meanings in some of mathematical literature, as explained on History of the separation axioms. Many of the concepts also have several names; the one listed first is preferred in Wikipedia.
Most of these axioms have alternative definitions with the same meaning; the definitions given here are those which fall into a consistent pattern relating the various notions of separation defined in the previous section. Other possible definitions can be found in the individual articles.
In all of the following definitions, X is again a topological space, and all functions are supposed to be continuous.
X is T_{0}, or Kolmogorov, if any two distinct points in X are topologically distinguishable. It will be a common theme among the separation axioms to have one version of an axiom that requires T_{0} and one version that doesn't.
X is R_{0}, or symmetric, if any two topologically distinguishable points in X are separated.
X is T_{1}, or accessible or Fréchet, if any two distinct points in X are separated. Thus, X is T_{1} if and only if it is both T_{0} and R_{0}. Although you may say such things as "T_{1} space", "Fréchet topology", and "Suppose that the topological space X is Fréchet", avoid saying "Fréchet space" in this context, since there is another entirely different notion of Fréchet space in functional analysis.
X is preregular, or R_{1}, if any two topologically distinguishable points in X are separated by neighbourhoods. Note that an R_{1} space must also be R_{0}.
X is Hausdorff, or T_{2} or separated, if any two distinct points in X are separated by neighbourhoods. Thus, X is Hausdorff if and only if it is both T_{0} and R_{1}. Note that a Hausdorff space must also be T_{1}.
X is T_{2½}, or Urysohn, if any two distinct points in X are separated by closed neighbourhoods. Note that a T_{2½} space must also be Hausdorff.
X is completely Hausdorff, or completely T_{2}, if any two distinct points in X are separated by a function. Note that a completely Hausdorff space must also be T_{2½}.
X is regular if, given any point x and closed set F in X, if x does not belong to F, then they are separated by neighbourhoods. In fact, in a regular space, any such x and F will also be separated by closed neighbourhoods. Note that a regular space must also be R_{1}.
X is regular Hausdorff, or T_{3}, if it is both T_{0} and regular. Note that a regular Hausdorff space must also be T_{2½}.
X is completely regular if, given any point x and closed set F in X, if x does not belong to F, then they are separated by a function. Note that a completely regular space must also be regular.
X is Tychonoff, or T_{3½}, completely T_{3}, or completely regular Hausdorff, if it is both T_{0} and completely regular. Note that a Tychonoff space must also be both regular Hausdorff and completely Hausdorff.
X is normal if any two disjoint closed subsets of X are separated by neighbourhoods. In fact, in a normal space, any two disjoint sets will also be separated by a function; this is Urysohn's lemma.
X is normal Hausdorff, or T_{4}, if it is both T_{1} and normal. Note that a normal Hausdorff space must also be both Tychonoff and normal regular.
X is completely normal if any two separated sets are separated by neighbourhoods. Note that a completely normal space must also be normal.
X is completely normal Hausdorff, or T_{5} or completely T_{4}, if it is both completely normal and T_{1}. Note that a T_{5} space must also be T_{4}.
X is perfectly normal if any two disjoint closed sets are precisely separated by a function. A perfectly normal space must also be completely normal.
X is perfectly normal Hausdorff, or perfectly T_{4}, if it is both perfectly normal and T_{1}. A perfectly T_{4} space must also be T_{5}.
Relationships between the axioms
The T_{0} axiom is special in that it cannot only be added to a property (so that regular plus T_{0} is T_{3}) but also subtracted from a property (so that Hausdorff minus T_{0} is preregular), in a fairly precise sense; see Kolmogorov quotient for more information. When applied to the separation axioms, this leads to the relationships in the table below:
T_{0} version  NonT_{0} version 

T_{0}  No requirement 
T_{1}  R_{0} 
Hausdorff  Preregular 
T_{2½}  No special name 
Completely Hausdorff  No special name 
Regular Hausdorff (T_{3})  Regular 
Tychonoff (T_{3½})  Completely regular 
Normal T_{0}  Normal 
Normal Hausdorff (T_{4})  Normal regular 
Completely normal T_{0}  Completely normal 
Completely normal Hausdorff (T_{5})  Completely normal regular 
Perfectly normal T_{0}  Perfectly normal 
Perfectly normal Hausdorff  Perfectly normal regular 
In this table, you go from the right side to the left side by adding the requirement of T_{0}, and you go from the left side to the right side by removing that requirement, using the Kolmogorov quotient operation.
Other than the inclusion or exclusion of T_{0}, the relationships between the separation axioms are indicated in the following diagram:
Separation_axioms.png
Hasse diagram of the separation axioms.
In this diagram, the nonT_{0} version of a condition is on the left side of the slash, and the T_{0} version is on the right side. Letters are used for abbreviation as follows: "P" = "perfectly", "C" = "completely", "N" = "normal", and "R" (without a subscript) = "regular". A bullet indicates that there is no special name for a space at that spot. The dash at the bottom indicates no condition.
You can combine two properties using this diagram by following the diagram upwards until both branches meet. For example, if a space is both completely normal ("CN") and completely Hausdorff ("CT_{2}"), then following both branches up, you find the spot "•/T_{5}". Since completely Hausdorff spaces are T_{0} (even though completely normal spaces may not be), you take the T_{0} side of the slash, so a completely normal completely Hausdorff space is the same as a T_{5} space.
As you can see from the diagram, normal and R_{0} together imply a host of other properties. Since regularity is the most well known of these, spaces that are both normal and R_{0} are typically called "normal regular spaces". In a somewhat similar fashion, T_{4} spaces are often called "normal Hausdorff spaces" by people that wish to avoid the "T" notation. (Wikipedia, in particular, wishes to avoid this notation, because it is less likely to be unambiguously understood.) These conventions can be generalised to other regular and Hausdorff spaces.
Other separation axioms
There are some other conditions on topological spaces that are sometimes classified with the separation axioms, but these don't fit in with the usual separation axioms as completely. Other than their definitions, they won't be discussed here.
X is semiregular if the regular open sets form a base for the open sets of X. Any regular space must also be semiregular.
X is fully normal if every open cover has an open star refinement. Every fully normal space must also be both normal regular and paracompact. In fact, fully normal spaces actually have more to do with paracompactness than with the usual separation axioms.
X is fully T_{4}, or fully normal Hausdorff, if it is both T_{1} and fully normal. Note that a fully T_{4} space must also be T_{4}.
Sources
 Schechter, Eric; 1997; Handbook of Analysis and its Foundations; http://www.math.vanderbilt.edu/~schectex/ccc/
 has R_{i} axioms (among others)
 Willard, Stephen; General Topology; AddisonWesley
 has all of the nonRi axioms mentioned in this article, with these definitions
 There are several other good books on general topology, but beware that some use slightly different definitions.de:Trennungsaxiom