Chern Simons Theory From Anyons

We could derive the Chern-Simons theory naturally from the study of anyons with fractional statistics in two-dimensional space. The time evolution of such anyon system could be expressed as a Berry phase, which is closely related to the Knizhnik-Zamolodchikov connection arise from conformal field theory.

As an introduction, the discussion below is intuitive and informal. However, since this story is quite simple (but deep), it is easy to make everything rigorous and precise.

Anyons with Fractional Statistics

It is well-known that in three (or higher) dimensional space, there are two types of fundamental particles: bosons and fermions, which differ by their exchange statistics. However, in two-dimensional space, there exists a more exotic type of particles called anyons, which exhibit fractional statistics that interpolate between bosonic and fermionic behavior. To be briefly, when two anyons are exchanged, the wavefunction acquires a phase factor 𝑒𝑖𝜃, where 𝜃 can take any value between 0 (bosons) and 𝜋 (fermions).

We can consider a system with two large mass anyons and statistic phase 𝜃. Thus, on the (for example) complex plane , the effective Lagrangian of this system could be written as:

𝐿𝜃𝑖=12𝑝𝑧𝑖𝑧̇𝑖+𝑖=12𝐴𝑧𝑖𝑧̇𝑖=𝑖=12𝑝𝑧𝑖𝑧̇𝑖+𝜃2𝜋𝑖𝑧̇1𝑧̇2𝑧1𝑧2,

where 𝑧1,𝑧2 denote the positions of the two anyons respectively. Due to the residue theorem., the holonomy of this gauge connection 𝐴𝑑𝑧𝑧 around one anyon would yield the desired phase factor 𝑒𝑖𝜃.

Now we consider the quantization of this system. Note that we have already fixed a gauge (the holomorphic gauge) for this gauge connection 𝐴, and the original theory has a gauge redundancy. Thus, the associated Schrödinger equation could arise from the constraint equation:

(𝑝𝑧𝑖+𝐴𝑧𝑖)𝜓(𝑧1,𝑧2)=0,𝑝𝑧𝑖𝜓(𝑧1,𝑧2)=0,𝑖=1,2.

Performing the canonical quantization 𝑝𝑧𝑖=𝑖𝜕𝑧𝑖, and switch to a new coordinate 𝑧=𝑧1𝑧2, such equations could be written as:

(𝜕𝑧𝜃2𝜋1𝑧)𝜓(𝑧,𝑧)=0,𝜕𝑧𝜓(𝑧,𝑧)=0.

The second equation implies that 𝜓 is holomorphic in 𝑧, and the first equation could be solved directly, yielding:

𝜓(𝑧1,𝑧2)=𝐶(𝑧1𝑧2)𝜃2𝜋,

which is a multi-valued function on with a branch cut at 𝑧1=𝑧2, and the monodromy around this branch cut is exactly the desired phase factor 𝑒𝑖𝜃.

A natural assumption is assume that such wave function would return to itself while winding 𝑘 times. Thus, the statistic phase 𝜃 is quantized as 𝜃=2𝜋𝑘, where 𝑘>0 (the charges 𝑞 would be absorbed into 𝑈(1) representation), which make the wave function become:

𝜓(𝑧1,𝑧2)=𝐶(𝑧1𝑧2)1𝑘.

Such a equation is exactly the famous (Abelian, two bodies) Knizhnik-Zamolodchikov equation arise from conformal field theory! And the associated KZ connection could be interpreted as the gauge connection 𝐴 we introduced above, which describes the statistics interaction between anyons.

While generalizing to 𝑛 anyons, the gauge connection could be expressed as:

𝐴𝑧𝑖=1𝑘𝑗𝑖Ω𝑖𝑗𝑧𝑖𝑧𝑗,

where Ω𝑖𝑗=𝑞𝑖𝑞𝑗, and above recipe could be repeated straightforwardly, and finally yield the Knizhnik-Zamolodchikov equation for 𝑛 bodies.

Time Evaluation

Now we consider the time evolution of such anyon system. Note that the Hamiltonian of this system is trivial 𝐻=0, thus, the time evolution would be fully determined by the Berry phase arise from the gauge connection 𝐴.

Thus, the only thing we need is to consider 𝑈(𝑡𝑏,𝑡𝑎)𝒯︀exp(𝑖𝑧(𝐴𝑧)𝑑𝑡), where 𝑧:𝑡2 is the world line of the anyon, and 𝒯︀exp denotes the time-ordered exponential. Such time evolution operator could be expanded as a Dyson series:

𝑈(𝑡𝑏,𝑡𝑎)=1+𝑛=1𝑖𝑛1𝑛!𝑡𝑎𝑡1𝑡𝑛𝑡𝑏𝐴𝑧(𝑧(𝑡1))𝐴𝑧(𝑧(𝑡𝑛))𝑑𝑡1𝑑𝑡𝑛1+𝑛=1(𝑖𝑘)𝑛𝒜︀𝑛.

Plugging the definition of the gauge connection 𝐴𝑧=1𝑘𝑖<𝑗1𝑧𝑖𝑧𝑗 into the above equation, we could rewrite the time evolution operator as:

𝒜︀𝑛=𝑃𝑙𝑃𝑡𝑎𝑡1𝑡𝑛𝑡𝑏𝑑𝑧𝑙1𝑑𝑧𝑙2𝑧𝑙1𝑧𝑙2Ω𝑙,

where 𝑃 is a pairing of the set {1,,𝑛}, each pair 𝑙𝑃 consists of two elements (𝑙1,𝑙2), and Ω𝑙=𝑞𝑙1𝑞𝑙2 is an double insertion of charges at the points 𝑧(𝑡𝑙1) and 𝑧(𝑡𝑙2) on the world line, could be written as Ω𝑖𝑗𝑞𝑖𝑞𝑗, which in fact, labels the 𝑈(1) representation.

Such expression (in non-Abelian case with some proper regularization) is Kontsevich’s integral formula for the link invariant, which is the Taylor coefficient of some link polynomial (for example, the Jones polynomial), called the universal Vassiliev invariant.

Abelian Chern-Simons Theory

Now we consider the field theory description of such anyon system. The effective field theory that describes the low-energy behavior of anyons is the Chern-Simons theory, which is a topological quantum field theory defined in three-dimensional spacetime. The Abelian Chern-Simons action is given by:

CS𝑋[𝐴]=𝑘4𝜋𝑋𝐴𝑑𝐴,

under the holomorphic gauge 𝐴=𝐴𝑡𝑑𝑡+𝐴𝑧𝑑𝑧, the Chern-Simons action reduces to:

CS𝑋[𝐴]𝑘4𝜋𝑋𝐴𝜕𝑡𝐴,

where 𝜕𝑡𝑑𝑡𝜕. And the anyon Lagrangian would be introduced into this action with minimal coupling, which could be expressed as:

𝑆anyon[𝐴,𝑧]=[𝐴𝑧(𝑧(𝑡))𝑧̇(𝑡)+𝐴0(𝑧(𝑡))]𝑑𝑡,

which is essentially the Wilson line observable in Chern-Simons theory, where the time is a natural parameter along the world line of the anyon.

In order to obtain the effective action of this theory, we need to integrate out the gauge field 𝐴. Since the Chern-Simons action is quadratic in 𝐴, such integration could be performed exactly. Note that, under the holomorphic gauge, correlation function (propagator) of the gauge field 𝐴 could be computed as:

𝐴(𝑧1,𝑡1)𝐴(𝑧2,𝑡2)=1𝑖𝑘𝑑𝑧1𝑑𝑧2𝑧1𝑧2𝛿(𝑡1𝑡2)𝑑𝑡1.

After performing the Gaussian integration, the effective action could be expressed as:

𝑆eff[𝑧]=1𝑘𝑑𝑧(𝑡)𝑑𝑧(𝑡)𝑧(𝑡)𝑧(𝑡)(𝑧̇(𝑡)𝑧̇(𝑡))𝑑𝑡=𝐿2𝜋/𝑘𝑑𝑡,

which is exactly the anyon Lagrangian we constructed above! Thus, we can conclude that, the Chern-Simons theory provides a natural field theory description of anyons with fractional statistics we constructed above.

Some Words about CS-WZW Correspondence

Though we won’t discuss the non-perturbative Chern-Simons theory in detail, it is still worth mentioning some important results from Witten’s work.

We’ve already know that, the wave function of the anyon system satisfies the Knizhnik-Zamolodchikov equation. Such equation arises naturally in the Wess-Zumino-Witten (WZW) model, which is the equation for the conformal blocks (which could be interpreted as local correlation functions) of the WZW model.

On the other hand, the wave function of Chern-Simons theory could be interpreted as the anyon wave function we constructed above. Thus, the CS-WZW correspondence could be interpreted as

ΨCS[𝑋,{𝐾𝑖,𝜌𝑖}]𝑖𝑉𝐾𝑖(𝜌𝑖)WZW,𝜕𝑋,

where 𝑉𝐾(𝜌) denotes the vertex operator insertion at the point where the line 𝐾 intersects the boundary 𝜕𝑋 in the WZW model, and {𝐾,𝜌} denotes a Wilson line 𝐾 colored by the representation 𝜌 of the gauge group. Here 𝜌 plays the role of the charge 𝑞 we introduced above.

Back to all articles

Loading comments...