In differential geometry and mathematical physics, a spin connection is a connection on a spinor bundle. It is induced, in a canonical manner, from the affine connection. It can also be regarded as the gauge field generated by local Lorentz transformations. In some canonical formulations of general relativity, a spin connection is defined on spatial slices and can also be regarded as the gauge field generated by local rotations.

The spin connection occurs in two common forms: the Levi-Civita spin connection, when it is derived from the Levi-Civita connection, and the affine spin connection, when it is obtained from the affine connection. The difference between the two of these is that the Levi-Civita connection is by definition the unique torsion-free connection, whereas the affine connection (and so the affine spin connection) may contain torsion.

Definition

Let ${\displaystyle e_{\mu }^{\;\,a))$ be the local Lorentz frame fields or vierbein (also known as a tetrad), which is a set of orthonormal space time vector fields that diagonalize the metric tensor

$${\displaystyle g_{\mu \nu }=e_{\mu }^{\;\,a}e_{\nu }^{\;\,b}\eta _{ab},}$$

${\displaystyle g_{\mu \nu ))$

${\displaystyle \eta _{ab))$

${\displaystyle g_{\mu \nu ))$

${\displaystyle e_{\mu }^{\;\,a))$

${\displaystyle g^{\mu \nu ))$

${\displaystyle g_{\mu \nu ))$

${\displaystyle \eta ^{ab))$

${\displaystyle \eta _{ab))$

${\displaystyle e^{\mu a}=g^{\mu \nu }e_{\nu }^{\;\,a))$

${\displaystyle e_{\nu a}=\eta _{ab}e_{\nu }^{\;\,b))$

The torsion-free spin connection is given by

$${\displaystyle \omega _{\mu }^{\ ab}=e_{\nu }^{\ a}\Gamma _{\ \sigma \mu }^{\nu }e^{\sigma b}+e_{\nu }^{\ a}\partial _{\mu }e^{\nu b}=e_{\nu }^{\ a}\Gamma _{\ \sigma \mu }^{\nu }e^{\sigma b}-e^{\nu b}\partial _{\mu }e_{\nu }^{\ a},}$$

${\displaystyle \Gamma _{\mu \nu }^{\sigma ))$

Note that ${\displaystyle \omega _{\mu }^{\ ab}=e_{\nu }^{\ a}\partial _{;\mu }e^{\nu b}=e_{\nu }^{\ a}(\partial _{\mu }e^{\nu b}+\Gamma _{\ \sigma \mu }^{\nu }e^{\sigma b})}$ using the gravitational covariant derivative ${\displaystyle \partial _{;\mu }e^{\nu b))$ of the contravariant vector ${\displaystyle e^{\nu b))$. The spin connection may be written purely in terms of the vierbein field as^[1]

$${\displaystyle \omega _{\mu }^{\ ab}={\tfrac {1}{2))e^{\nu a}(\partial _{\mu }e_{\nu }^{\ b}-\partial _{\nu }e_{\mu }^{\ b})-{\tfrac {1}{2))e^{\nu b}(\partial _{\mu }e_{\nu }^{\ a}-\partial _{\nu }e_{\mu }^{\ a})-{\tfrac {1}{2))e^{\rho a}e^{\sigma b}(\partial _{\rho }e_{\sigma c}-\partial _{\sigma }e_{\rho c})e_{\mu }^{\ c},}$$

${\displaystyle a,b}$

The spin connection ${\displaystyle \omega _{\mu }^{\ ab))$ defines a covariant derivative ${\displaystyle D_{\mu ))$ on generalized tensors. For example, its action on ${\displaystyle V_{\nu }^{\ a))$ is

$${\displaystyle D_{\mu }V_{\nu }^{\ a}=\partial _{\mu }V_{\nu }^{\ a}+((\omega _{\mu ))^{a))_{b}V_{\nu }^{\ b}-\Gamma _{\ \nu \mu }^{\sigma }V_{\sigma }^{\ a))$$

Cartan's structure equations

In the Cartan formalism, the spin connection is used to define both torsion and curvature. These are easiest to read by working with differential forms, as this hides some of the profusion of indexes. The equations presented here are effectively a restatement of those that can be found in the article on the connection form and the curvature form. The primary difference is that these retain the indexes on the vierbein, instead of completely hiding them. More narrowly, the Cartan formalism is to be interpreted in its historical setting, as a generalization of the idea of an affine connection to a homogeneous space; it is not yet as general as the idea of a principal connection on a fiber bundle. It serves as a suitable half-way point between the narrower setting in Riemannian geometry and the fully abstract fiber bundle setting, thus emphasizing the similarity to gauge theory. Note that Cartan's structure equations, as expressed here, have a direct analog: the Maurer–Cartan equations for Lie groups (that is, they are the same equations, but in a different setting and notation).

Writing the vierbeins as differential forms

$${\displaystyle e^{a}=e_{\mu }^{\;\,a}dx^{\mu ))$$

$${\displaystyle \omega ^{ab}=\omega _{\mu }^{\;\;ab}dx^{\mu ))$$

$${\displaystyle \Theta ^{a}=de^{a}+\omega _{\;b}^{a}\wedge e^{b))$$

$${\displaystyle R_{\;\,b}^{a}=d\omega _{\;\,b}^{a}+\omega _{\;c}^{a}\wedge \omega _{\;\,b}^{c}={\tfrac {1}{2))R_{\;\,bcd}^{a}e^{c}\wedge e^{d))$$

$${\displaystyle d\Theta ^{a}+\omega _{\;b}^{a}\wedge \Theta ^{b}=R_{\;\,b}^{a}\wedge e^{b))$$

$${\displaystyle dR_{\;\,b}^{a}+\omega _{\;c}^{a}\wedge R_{\;\,b}^{c}-R_{\;\,c}^{a}\wedge \omega _{\;b}^{c}=0.}$$

${\displaystyle V_{\;\,b}^{a))$

$${\displaystyle DV_{\;\,b}^{a}=dV_{\;\,b}^{a}+\omega _{\;c}^{a}\wedge V_{\;\,b}^{c}-(-1)^{p}V_{\;\,c}^{a}\wedge \omega _{\;b}^{c}.}$$

$${\displaystyle DR_{\;\,b}^{a}=0.}$$

Derivation

Metricity

It is easy to deduce by raising and lowering indices as needed that the frame fields defined by ${\displaystyle g_{\mu \nu }={e_{\mu ))^{a}{e_{\nu ))^{b}\eta _{ab))$ will also satisfy ${\displaystyle {e_{\mu ))^{a}{e^{\mu ))_{b}=\delta _{b}^{a))$ and ${\displaystyle {e_{\mu ))^{b}{e^{\nu ))_{b}=\delta _{\mu }^{\nu ))$. We expect that ${\displaystyle D_{\mu ))$ will also annihilate the Minkowski metric ${\displaystyle \eta _{ab))$,

$${\displaystyle D_{\mu }\eta _{ab}=\partial _{\mu }\eta _{ab}-{\omega _{\mu a))^{c}\eta _{cb}-{\omega _{\mu b))^{c}\eta _{ac}=0.}$$

${\displaystyle {\omega _{\mu ))^{ab}=-{\omega _{\mu ))^{ba}.}$

${\displaystyle \partial _{;\beta }({e_{\mu ))^{a}{e^{\mu ))_{b})=0}$

${\displaystyle \partial _{;\beta }{e_{\mu ))^{a}{e^{\mu ))_{b}=-{e_{\mu ))^{a}\partial _{;\beta }{e^{\mu ))_{b))$

${\displaystyle {\omega _{\beta ))^{ab}=-{\omega _{\beta ))^{ba))$

${\displaystyle g_{\mu \nu ;\alpha }=0.}$

By substituting the formula for the Christoffel symbols ${\displaystyle {\Gamma ^{\nu ))_{\sigma \mu }={\tfrac {1}{2))g^{\nu \delta }\left(\partial _{\sigma }g_{\delta \mu }+\partial _{\mu }g_{\sigma \delta }-\partial _{\delta }g_{\sigma \mu }\right)}$ written in terms of the ${\displaystyle {e_{\mu ))^{a))$, the spin connection can be written entirely in terms of the ${\displaystyle {e_{\mu ))^{a))$,

$${\displaystyle {\omega _{\mu ))^{ab}=e^{\nu [a}(((e_{\nu ))^{b]))_{,\mu }-((e_{\mu ))^{b]))_{,\nu }+e^{\sigma |b]}{e_{\mu ))^{c}e_{\nu c,\sigma })}$$

By the metric compatibility

This formula can be derived in another way. To directly solve the compatibility condition for the spin connection ${\displaystyle {\omega _{\mu ))^{ab))$, one can use the same trick that was used to solve ${\displaystyle \nabla _{\rho }g_{\alpha \beta }=0}$ for the Christoffel symbols ${\displaystyle {\Gamma ^{\gamma ))_{\alpha \beta ))$. First contract the compatibility condition to give

$${\displaystyle {e^{\alpha ))_{b}{e^{\beta ))_{c}(\partial _{[\alpha }e_{\beta ]a}+{\omega _{[\alpha a))^{d}\;e_{\beta ]d})=0.}$$

Then, do a cyclic permutation of the free indices ${\displaystyle a,b,}$ and ${\displaystyle c}$, and add and subtract the three resulting equations:

$${\displaystyle \Omega _{bca}+\Omega _{abc}-\Omega _{cab}+2{e^{\alpha ))_{b}\omega _{\alpha ac}=0}$$

${\displaystyle \Omega _{bca}:={e^{\alpha ))_{b}{e^{\beta ))_{c}\partial _{[\alpha }e_{\beta ]a))$

$${\displaystyle \omega _{\alpha ca}={\tfrac {1}{2)){e_{\alpha ))^{b}(\Omega _{bca}+\Omega _{abc}-\Omega _{cab}).}$$

From this we obtain the same formula as before.

Applications

The spin connection arises in the Dirac equation when expressed in the language of curved spacetime, see Dirac equation in curved spacetime. Specifically there are problems coupling gravity to spinor fields: there are no finite-dimensional spinor representations of the general covariance group. However, there are of course spinorial representations of the Lorentz group. This fact is utilized by employing tetrad fields describing a flat tangent space at every point of spacetime. The Dirac matrices ${\displaystyle \gamma ^{a))$ are contracted onto vierbiens,

$${\displaystyle \gamma ^{a}{e^{\mu ))_{a}(x)=\gamma ^{\mu }(x).}$$

We wish to construct a generally covariant Dirac equation. Under a flat tangent space Lorentz transformation the spinor transforms as

$${\displaystyle \psi \mapsto e^{i\epsilon ^{ab}(x)\sigma _{ab))\psi }$$

We have introduced local Lorentz transformations on flat tangent space generated by the ${\displaystyle \sigma _{ab))$'s, such that ${\displaystyle \epsilon _{ab))$ is a function of space-time. This means that the partial derivative of a spinor is no longer a genuine tensor. As usual, one introduces a connection field ${\displaystyle {\omega _{\mu ))^{ab))$ that allows us to gauge the Lorentz group. The covariant derivative defined with the spin connection is,

$${\displaystyle \nabla _{\mu }\psi =\left(\partial _{\mu }-{\tfrac {i}{4)){\omega _{\mu ))^{ab}\sigma _{ab}\right)\psi =\left(\partial _{\mu }-{\tfrac {i}{4))e^{\nu a}\partial _{;\mu }{e_{\nu ))^{b}\sigma _{ab}\right)\psi ,}$$

$${\displaystyle (i\gamma ^{\mu }\nabla _{\mu }-m)\psi =0.}$$

The generally covariant fermion action couples fermions to gravity when added to the first order tetradic Palatini action,

$${\displaystyle {\mathcal {L))=-{1 \over 2\kappa ^{2))e\,{e^{\mu ))_{a}{e^{\nu ))_{b}{\Omega _{\mu \nu ))^{ab}[\omega ]+e{\overline {\psi ))(i\gamma ^{\mu }\nabla _{\mu }-m)\psi }$$

${\textstyle e:=\det {e_{\mu ))^{a}={\sqrt {-g))}$

${\displaystyle {\Omega _{\mu \nu ))^{ab))$

The tetradic Palatini formulation of general relativity which is a first order formulation of the Einstein–Hilbert action where the tetrad and the spin connection are the basic independent variables. In the 3+1 version of Palatini formulation, the information about the spatial metric, ${\displaystyle q_{ab}(x)}$, is encoded in the triad ${\displaystyle e_{a}^{i))$ (three-dimensional, spatial version of the tetrad). Here we extend the metric compatibility condition ${\displaystyle D_{a}q_{bc}=0}$ to ${\displaystyle e_{a}^{i))$, that is, ${\displaystyle D_{a}e_{b}^{i}=0}$ and we obtain a formula similar to the one given above but for the spatial spin connection ${\displaystyle \Gamma _{a}^{ij))$.

The spatial spin connection appears in the definition of Ashtekar–Barbero variables which allows 3+1 general relativity to be rewritten as a special type of ${\displaystyle \mathrm {SU} (2)}$ Yang–Mills gauge theory. One defines ${\displaystyle \Gamma _{a}^{i}=\epsilon ^{ijk}\Gamma _{a}^{jk))$. The Ashtekar–Barbero connection variable is then defined as ${\displaystyle A_{a}^{i}=\Gamma _{a}^{i}+\beta c_{a}^{i))$ where ${\displaystyle c_{a}^{i}=c_{ab}e^{bi))$ and ${\displaystyle c_{ab))$ is the extrinsic curvature and ${\displaystyle \beta }$ is the Immirzi parameter. With ${\displaystyle A_{a}^{i))$ as the configuration variable, the conjugate momentum is the densitized triad ${\displaystyle E_{a}^{i}=\left|\det(e)\right|e_{a}^{i))$. With 3+1 general relativity rewritten as a special type of ${\displaystyle \mathrm {SU} (2)}$ Yang–Mills gauge theory, it allows the importation of non-perturbative techniques used in Quantum chromodynamics to canonical quantum general relativity.