International Journal on Future Revolution in Computer Science & Communication Engineering ISSN: 2454-4248

Volume: 4 Issue: 9 12 – 20
_______________________________________________________________________________________________
Non-Abelian Transformations in SU (2), SU (3) and Discrete Symmetries
Dr. Dev Raj Mishra
Department of Physics,
R.H.Government Post Graduate College, Kashipur, U.S.Nagar,
Uttarakhand -244713 INDIA
e-mail: dr_devraj_mishra@yahoo.co.in

Abstract — The pedagogical introduction about the non-abelian transformations and their application in SU(2), SU(3) symmetries is presented.
Chiral Gauge groups and discrete symmetries like parity and charge congjugation are also discussed.

Keywords-Non-Abelian transformations,Gauge field,QCD, discrete symmetries

__________________________________________________*****_________________________________________________
The matrices U can be written in exponentialform
I. INTRODUCTION 𝑈 = exp⁡(𝜉 𝑎 𝑡𝑎 ) (3)
The abelian transformations as for electrodynamics where the matrices 𝑡 𝑎 are the generators of the representation of
a
depend on a single parameter ξ. Theyare invariant under local the group and the ξ constitute a set of linearly independentreal
phase transformationsandare commuting. That means if the parameters in terms of which the group elements are described.
order of the transformation is changed, outcome is the same. The dimension of the group is the number of
Thenon-abelian transformations, on the other hand, depend on generators,which is obviously equal to the number
a
several parameters and do not necessarily commute. Therefore ofindependent parameters ξ .However, dimension of the group
the outcome of twoconsecutive transformations depends on the is unrelatedto the dimension of the matrices U and ta. It is often
orderin which the transformations are performed.The straightforwardto determine the dimension or the number of
generators of these transformations, as we shall see below, also generators of a group. For example, the generators of the
do not commute. SO(N) group areN × N real and anti-symmetricmatrices, in
Further the number of parameters, on which anon-abelian order that Eq.(1) defines an orthogonal matrix,UT= U−1.
gauge transformationdepends, indicates the number of Therefore total number of elements possible is N 2 − N as N is
independent gauge fields.Thereforethe gauge fields constitute the diagonal elements. Due to condition of anti-symmetricity,
the so-called representation group, whose member 1
𝑁(𝑁 − 1) independent matrices are possible. Therefore the
elementstransform among them. This is similar to the 2
1
componentsof a three-dimensional vector which transform dimensionof the SO(N) group is equal to 𝑁(𝑁 − 1) . The
2
among themselves under thegroup of rotations. These gauge SU(N) group with the defining relation U†= U−1 requires the
fields are self-interacting as the matricesrepresenting these a
generators t to be anti-hermiteanN×Nmatrices. Furthermore,
fields do not commute.Let some fields are represented by member matrices U should haveunit determinant. As we know,
matrices U which are elements of a group G.These matrices are a
expressible in terms of generators 𝑡𝑎 which are said to obey the anti-hermitean matrices t are traceless matrices.
2
𝑐 𝑐
following equation, 𝑡𝑎 , 𝑡𝑏 = 𝑓𝑎𝑏 𝑡𝑐 withnon-zero 𝑓𝑎𝑏 .These ThereforeN −1 independent anti-hermitean traceless matrices
commutationrelations are known as the Lie algebra and the are possible and so the dimensionoftheSU(N) group is equal to
2
corresponding group G which follows this algebra is called a N − 1.
𝑐 Photon or more precisely the photon field is electrically
Lie group. The proportionality constants 𝑓𝑎𝑏 are called the
neutral and therefore does not couple to itself. However, these
structure constants,because they define the multiplication
non-commuting gauge fieldsare not electrically neutral and
properties of the Lie group G. The expression 𝑡𝑎 , 𝑡𝑏 ≠ 0
interact with each other. Here we shall discuss the invariance of
indicates the non abelian nature of the group elements(Please
Lagrangian under non-abelian transformations.
see Ref.[1-7] for detail).
Let us consider an infinitesimal transformations, where
a
II. NON-ABELIAN TRANSFORMATION the parameters ξ aresmallso that from Eq.(3), we have
The group transformation𝑈 transforms the fields 𝜓 𝑥 as 𝑈 = 1 + 𝜉𝑎 𝑡 𝑎 + 𝑂 𝜉2 . (4)
Because the matrices U defined in Eq.(1) constitute a
follows
representation of thegroup, these matrices must be closed under
𝜓 𝑥 → 𝜓 ′ 𝑥 = 𝑈𝜓(𝑥) (1)
multiplication and so should be expressible into the same
Here 𝜓 𝑥 in general, denotes an array of different field
exponentialform. This leads to an important condition on the
components written as a column vector. In terms of a
components, we may write matrices t that their commutator is a linear combination of the
𝜓𝑖 𝑥 → 𝜓𝑖′ 𝑥 = 𝑈𝑖𝑗 𝜓𝑗 (𝑥) (2) same set of generatorsi.e.
𝑐
𝑡𝑎 , 𝑡𝑏 = 𝑓𝑎𝑏 𝑡𝑐 (5)
12
IJFRCSCE | September 2018, Available @ http://www.ijfrcsce.org
_______________________________________________________________________________________
International Journal on Future Revolution in Computer Science & Communication Engineering ISSN: 2454-4248
Volume: 4 Issue: 9 12 – 20
_______________________________________________________________________________________________
Differentiating Eq.(1), we see transformations U1 and U2give U3 = U1U2,whereU3 is another
′
𝜕𝜇 𝜓 𝑥 → 𝜕𝜇 𝜓 𝑥 member of the group. To show this, we have on applying U1
(6) 𝑊𝜇 → 𝑊𝜇′′ = 𝑈1 𝑊𝜇 𝑈1−1 + 𝜕𝜇 𝑈1 𝑈1−1
= 𝑈 𝑥 𝜕𝜇 𝜓 𝑥 + 𝜕𝜇 𝑈 𝑥 𝜓 𝑥 Further applying U2 , we get
Therefore𝜓 𝑥 transform according to Eq.(1) i.e. 𝜓 𝑥 → 𝑊𝜇 → 𝑊𝜇′′ = 𝑈2 𝑈1 𝑊𝜇 𝑈1−1 + 𝜕𝜇 𝑈1 𝑈1−1 𝑈2−1 + 𝜕𝜇 𝑈2 𝑈2−1
𝜓 ′ 𝑥 = 𝑈𝜓(𝑥) and such quantities are called as covariant = 𝑈2 𝑈1 𝑊𝜇 𝑈2 𝑈1 −1 + 𝑈2 𝜕𝜇 𝑈1 + 𝜕𝜇 𝑈2 𝑈1 𝑈1−1 𝑈2−1
−1 −1
quanties. However, due to the presence of the second term on = 𝑈2 𝑈1 𝑊𝜇 𝑈2 𝑈1 + 𝜕𝜇 𝑈2 𝑈1 𝑈2 𝑈1 (14)
the right-hand side of Eq.(6), derivative 𝜕𝜇 𝜓(𝑥) does not The consistency of (13) requires that term
transform covariantly. Therefore ∂μneeds to be replaced by a ∂𝜇 𝑈 𝑈 −1 on the right-hand side should also satisfy Lie-
covariant quantity called covariant derivative Dμ. The algebracommutation rule. An infinitesimally small variation
transformation rule for Dμ should be 𝑑𝑥 𝜇 will change U by a small amount because U(x) isa
′
𝐷𝜇 𝜓 𝑥 → 𝐷𝜇 𝜓 𝑥 = 𝑈 𝑥 𝐷𝜇 𝜓 𝑥 (7) differentiable function of the space-time coordinates. Therefore
As we know for abelian transformations [see for example 𝑈 𝑥 𝜇 + 𝑑𝑥 𝜇 = 𝐼 + 𝑐𝜇𝑎 𝑥 𝑑𝑥𝑡𝑎 + ⋯ ⋯ 𝑈 𝑥 (15)
1, 2]. Also using Taylor expansion for the left-hand side, we
𝜓 𝑥 → 𝜓 ′ 𝑥 = 𝑒 𝑖𝑞𝜉 𝜓(𝑥) (8) get
Comparing with Eq.(1) we can say here the 𝑈 𝑥 𝜇 + 𝑑𝑥 𝜇 = 𝑈 𝑥 + 𝜕𝜇 𝑈 𝑥 𝑑𝑥 𝜇 + ⋯ ⋯ (16)
transformation 𝑈 = 𝑒 𝑖𝑞𝜉 . Although U is abelian here and is Therefore comparing 𝑑𝑥 𝜇 term in (15) and (16), we have
called the phase transformation which generates a group of
1 × 1 unitary matrices called U(1). It has only one parameter q 𝜕𝜇 U x U −1 = cμa t a (17)
which measures the strength of the transformation. The which is being proportional to the generator t a follows
covariant derivative for this abelian transformation is given by Lie-algebra. Further using Eq.(3) we have
𝐷𝜇 ≡ ∂𝜇 − 𝑖 qA 𝜇 (9) ∂μ U x = ∂μ ξa t a exp ξb t b = ∂μ ξa t a 𝑈
WhereA𝜇 follows a transformation rule of (18)
Or ∂μ U x 𝑈 −1 = ∂μ ξa t a
A𝜇 → A𝜇 = A𝜇 + 𝜕𝜇 𝜉 (10) The first term in Eq.(13) 𝑈W𝜇 𝑈 −1 follows lie algebra.
Where∂𝜇 is the ordinary derivative, is the called the This is because if a group element X, is sandwichedbetween the
coupling strength and A𝜇 is called the gauge field for the transformation matrix U and its inverse, is again an element
transformation. The infinitesimal displacement 𝜉 is called a ofthe group,
covariant translation. This formula for covariant derivative 𝐷𝜇 i. e. if 𝑋, 𝑈 ∈ G, Y = UXU −1 ∈ G. (19)
can be generalized for the covariant derivative for an arbitrary X, being member of G, we have
group. Namely, we takethelinear combination of an ordinary 𝑋 = 𝐼 + 𝜉 𝑎 𝑡𝑎 + 𝑂(𝜉 2 ),
derivative and an infinitesimal gaugetransformation, where the Also 𝑈 = exp 𝜉 𝑎 𝑡𝑎 ≈ 𝐼 + 𝜉 𝑎 𝑡𝑎 + 𝑂 𝜉 2 ,
parameters of the latter define the non-abelian gauge Let 𝑌 = UXU −1 = 𝐼 + 𝜉 𝑎 𝑡𝑎 + 𝑂(𝜉 2 ),
fields.Therefore, for a non-abelian gauge field W𝜇 , the Substituting this into Eq.(191919), We get
covariant derivative can be written as 𝑌 = 𝐼 + 𝜉 𝑎 𝑡𝑎 + 𝑂 𝜉 2
𝐷𝜇 𝜓 ≡ ∂𝜇 𝜓 − W𝜇 𝜓 (11) = 𝐼 + 𝜉 𝑎 𝑡𝑎 + 𝑂 𝜉 2 𝐼 + 𝜉 𝑎 𝑡𝑎 + 𝑂(𝜉 2 )
W𝜇 has the characteristic feature of a gauge field, as it 𝐼 − 𝜉 𝑎 𝑡𝑎 + 𝑂 𝜉 2
carries the information regarding the group from one space- = 𝐼 + 𝜉 𝑎 𝑡𝑎 + 𝑂 𝜉 2
time point to another. This exercise shows that 𝜉 𝑎 = 𝑂 𝜉𝑎 .Therefore, 𝜉 𝑎 can
Here W𝜇 is represented by a matrix which is a member also be expanded in a power series in 𝜉 𝑎 . Thereforethe term
of a Lie Group Gand therefore, can be expressed as a linear UWμ U −1 in (13) is also Lie-algebra valued and is called an
combination of its generators t 𝑎 . adjoint representation of Wμ . Therefore the conclusion is that
W𝜇 = W𝜇𝑎 𝑡𝑎 (12) the gauge fields 𝑊𝜇 transform in an adjoint representation of
Let us transform W𝜇 𝜓 under gauge transformation as the group UWμ U −1 , added with an inhomogeneous but Lie
defined by Eq.(6) as algebra valued term ∂𝜇 𝑈 𝑈 −1 . It is easy to evaluate
′ ′
W𝜇 𝜓 ≡ ∂𝜇 𝜓 ′ − 𝐷𝜇 𝜓 transformation rule for components 𝑊𝜇𝑎 by using Eq.(3),
= 𝑈 ∂𝜇 𝜓 + ∂𝜇 𝑈 𝜓 − 𝑈 𝐷𝜇 𝜓 Eq.(12)and Eq.(13)
′
= 𝑈W𝜇 𝑈 −1 + ∂𝜇 𝑈 𝑈 −1 𝜓 ′ 𝑊𝜇𝑎 → 𝑊𝜇𝑎 = 𝑊𝜇𝑎 + 𝑓𝑏𝑐𝑎 𝜉 𝑏 𝑊𝜇𝑐 + ∂μ ξa + 𝑂(ξ2 ) (20)
Therefore, thetransformation rule for Wμ is This transformation law differs from that of the abelian
W𝜇 → W𝜇′ ≡ 𝑈W𝜇 𝑈 −1 + ∂𝜇 𝑈 𝑈 −1 (13) gauge fields by thepresence of the term 𝑓𝑏𝑐𝑎 𝜉 𝑏 𝑊𝜇𝑐 . Eq. (20) can
Clearly the gauge field W does not transform covariantly also be written
but it defines a group. Successiveapplication of two δWμa = Dμ ξa + O(ξ2 ) (21)
13
IJFRCSCE | September 2018, Available @ http://www.ijfrcsce.org
_______________________________________________________________________________________
International Journal on Future Revolution in Computer Science & Communication Engineering ISSN: 2454-4248
Volume: 4 Issue: 9 12 – 20
_______________________________________________________________________________________________
a
Whereξ isa quantity which transformsin the adjoint Unlike abelian field strength Fμν , Gμν is not invariant
representation ofthe gauge group. HereDμis the generatorof under an infinitesimal transformation, it transforms as
′
covariant translations, which are infinitesimal space- Gμν → Gμν = Gμν + 𝜉 𝑎 t 𝑎 , Gμν (29)
timetranslations and infinitesimal field-dependent gauge Or in terms of components
𝑎 ′
transformations preserving the covariant nature of translated 𝑎
𝐺𝜇𝜈 = 𝐺𝜇𝜈 𝑎
= 𝐺𝜇𝜈 + 𝑓𝑏𝑐𝑎 𝜉 𝑏 Gμν
c (30)
quantity. This is equally true for infinitesimal abelian and non- The adjoint representation is trivial for an abelian group,
abelian transformations. therefore the structure constants 𝑓𝑏𝑐𝑎 vanish for the abelian
The transformations satisfy δ(φψ) = (δφ)ψ+φ(δψ), transformations and the field strengths are invariant.
Therefore the Leibniz’ rule is applicable tocovariant Using Eq. (24), we can write
derivativesi.e. [𝐷𝜇 , 𝐷𝜈 ] = −𝐺𝜇𝜈 (31)
Dμ (φψ) = (Dμ φ)ψ + φ Dμ ψ (22) This result is known as the Ricci identity. This implies that
The covariant derivatives depend on the representation of the commutator of two covariant derivatives for a non-abelian
the fieldsonwhich they act through the choice of the generators field is equal to an infinitesimal gauge transformation with
a 𝑎
t . Hence each of the threeterms in Eq.(22) may contain a parameters−𝐺𝜇𝜈 .
different representation for the generators. Further, as for the abelian case, using Eq.(31) and
Let us understand this from a simpler example for abelian applying Jacobi identity, we get the cyclic combination of three
transformations. Let 𝜓1 and 𝜓2 transform under local phase covariant derivatives sum up to zero, that is
transformations with strength q1 and q2 respectively, then we 𝐷𝜇 𝐷𝜈 , 𝐷𝜌 + 𝐷𝜈 𝐷𝜌 , 𝐷𝜇 + 𝐷𝜌 𝐷𝜇 , 𝐷𝜈 = 0
have
Or 𝐷𝜇 𝐺𝜈𝜌 + 𝐷𝜈 𝐺𝜌𝜇 + 𝐷𝜌 𝐺𝜇𝜈 = 0 (32)
𝐷𝜇 𝜓1 𝜓2 = 𝜕𝜇 − 𝑖 𝑞1 + 𝑞2 𝐴𝜇 𝜓1 𝜓2 ,
Since Dμ ≡ ∂μ − Wμ , it can be shown that
𝐷𝜇 𝜓1 = 𝜕𝜇 − 𝑖𝑞1 𝐴𝜇 𝜓1 , (23)
𝐷𝜇 𝐺𝜈𝜌 = ∂μ 𝐺𝜈𝜌 − 𝑊𝜇 , 𝐺𝜈𝜌 (33)
𝐷𝜇 𝜓2 = 𝜕𝜇 − 𝑖𝑞2 𝐴𝜇 𝜓2 ,
or, in components,
With these definitions it is straight forward to verify Eq.
(22).
𝑎
𝐷𝜇 𝐺𝜈𝜌 = ∂μ 𝐺𝜈𝜌 𝑎
− 𝑓𝑏𝑐𝑎 𝑊𝜇𝑏 𝐺𝜈𝜌𝑐
(34)
However, unlike ordinary differentiations, two covariant Eq.(32) is known as the Bianchi identity and in the abelian
differentiations for non abelian fieldsdo not commute. case it corresponds to the Homogeneous Maxwell equations. As
Thecommutator of two covariantderivatives DμandDνfor non weknow the integral form of the homogeneous Maxwell
abelian fields, which is obviously a covariant quantity, is given equation implies the conservationof magnetic flux. Thenon-
by abelian analogue of the magnetic flux is not conservedand is
D μ , Dν ψ = D μ D ν ψ − D ν D μ ψ carried away by the non-abelian gauge fields because of the
(24) presence of commutator term in Eq.(33) which is zero in the
= − ∂μ Wν − ∂ν Wμ − Wμ , Wν ψ case of an abelian transformation.This suggests the existence of
Thequantity in parenthesis is a covariant anti-symmetric a source of magnetic lines of force i.e. a magnetic monopole.
tensorGμν,
𝐺𝜇𝜈 = ∂μ Wν − ∂ν Wμ − Wμ , Wν (25) III. RESCALING OF THE GAUGE FIELD AND
which is called as the field strength. This quantity is analogous COUPLING CONSTANT
to abelian case field tensor𝐹𝜇𝜈 = ∂μ Aν − ∂ν Aμ . It differs from The coupling constant, g, rescales thegauge fields
the abelian field strength 𝐹𝜇𝜈 by the presence of the term WaμtogWaμ. If the gauge group is a product of
quadratic in Wμ .We saw by Eq. (13) that gauge field Wμ does differentsubgroups, then the gauge fields corresponding to each
not transform covariantly. The covariant object associated with one ofthese subgroups is rescaled by an independent coupling
the gauge field isjust the field strength 𝐺𝜇𝜈 . BothDνDμψand constant. The number of coupling constants isthus equal to the
DμDνψtransform covariantly under the gaugetransformations, number of independent gauge-invariant terms that
therefore the field strength must transform covariantly dependexclusively on the gauge fields.
accordingto Let us consider a simple gauge group which is defined as
Gμν → Gμν′
= UGμν U −1 (26) a group without invariantsubgroups. Therefore there is only one
coupling constant.Thequantities used in above discussion will
Further𝑊μ is Lie-algebra valued and the quadratic term in
require the following substitutions
(25) is a commutator,therefore field strength Gμνis also Lie-
𝑊𝜇𝑎 → 𝑔𝑊𝜇𝑎 , 𝑎
𝐺𝜇𝜈 𝑎
→ 𝑔𝐺𝜇𝜈 , ξ𝑎 → 𝑔ξ𝑎 (35)
algebra valued and is decomposed in terms of the group
𝐷𝜇 𝜓 ≡ ∂𝜇 𝜓 − gW𝜇 𝜓 (36)
generatorsta, mathematically
a The Ricci identity (31) now modifies to
Gμν = Gμν ta (27)
𝑎 [𝐷𝜇 , 𝐷𝜈 ] = −𝑔𝐺𝜇𝜈 (37)
wherethe matrices 𝐺𝜇𝜈 is shown as
and the field strength tensor
𝑎
𝐺𝜇𝜈 = ∂μ Wνa − ∂ν Wμa − 𝑓𝑏𝑐𝑎 Wμb Wνc (28)
𝐺𝜇𝜈 = ∂μ Wν − ∂ν Wμ − g Wμ , Wν (38)
14
IJFRCSCE | September 2018, Available @ http://www.ijfrcsce.org
_______________________________________________________________________________________
International Journal on Future Revolution in Computer Science & Communication Engineering ISSN: 2454-4248
Volume: 4 Issue: 9 12 – 20
_______________________________________________________________________________________________
The components Gμν a a
defined by the relation Gμν = Gμν ta 𝑈 𝜉 = 𝑒𝑥𝑝 𝑔𝜉 𝑎 𝑡𝑎 (a = 1,2,3) (50)
rescale to Here 𝜉 is the rotation or translational parameter in
𝐺𝜇𝜈𝑎
= ∂μ Wνa − ∂ν Wμa − 𝑔𝑓𝑏𝑐𝑎 Wμb Wνc (39) isospinspace.The coupling constant g is in accordance with
The gauge-transformation parameters are also rescaled, Eq.(40).The three generators of SU(2) are expressed in terms of
and therefore gauge transformation U becomes the isotopic spinmatrices𝜎a
𝑈 = exp⁡ (𝑔𝜉 𝑎 𝑡𝑎 ) (40) 1
𝑡𝑎 = 𝑖𝜏𝑎 , (51)
Therefore the field components 𝑊𝜇𝑎 , 𝐺𝜇𝜈 𝑎
and the matter 2
whichare the Pauli matrices used in the context of ordinary
field 𝜓 transform under infinitesimal gauge transformations as
spin,
𝛿Wμa = 𝜕𝜇 ξ𝑎 − 𝑔𝑓𝑏𝑐𝑎 Wμb ξc = 𝐷𝜇 ξ𝑎 (41) 0 1 0 −𝑖 1 0
𝑎 𝑎 b 𝑐
𝛿𝐺𝜇𝜈 = 𝑔𝑓𝑏𝑐 ξ 𝐺𝜇𝜈 𝜏1 = ,𝜏 = , 𝜏3 = (52)
(42) 1 0 2 𝑖 0 0 −1
𝛿𝜓 = 𝑔ξ𝑎 𝑡𝑎 𝜓 (43) It is easy to see that the generators tasatisfy the
The quantity ξawhilein rewriting the gauge field commutationrelations
transformation as a covariant derivative is treated like a [𝑡𝑎 , 𝑡𝑏 ] = −𝜀𝑎𝑏𝑐 𝑡𝑐 , (53)
quantity that transforms in the adjointrepresentation of the ensuring that the transformation matrices in Eq.(50) form a
group. group.
IV. GAUGE THEORY OF SU(2) Historically an interesting early attempt to construct the
The gauge-invariant Lagrangians for spin-0 and spin-1/2 covariant derivative on ψ was presented by O. Klein (1938) at
fields can now be constructed using covariant derivatives. Let the motivation of Yukawa’s conjecture of the existence of
us consider a set of N spinor fields ψitransforming under thepion. It was based on a somewhat ad-hoc, extension
ofgravity in five space-time dimensions, and led to theories
unitary transformations U i.e. U†= U−1.This belongs to a certain
which have features very similarto those of SU(2) and SU(2) ×
group G,we have
U(1) gauge theories. But the present day non-abelian gauge
𝜓𝑖 → 𝜓′𝑖 = 𝑈𝑖𝑗 𝜓𝑗 (44)
field theory was constructed by Yang and Mills and
(i, j = 1, . . . ,N). The indices i, j can be suppressed and ψ can be independently by Shaw [2]. It isbased on above discussed
considered as an N-dimensional column vector SU(2) matrices and is motivated by the existence of the
𝜓 → 𝜓′ = 𝑈𝜓 (45) approximate iso-spin invariance in Nature.The observation that
Similarly we can write for 𝜓 the proton and neutron are almost degenerate in mass, and play
𝜓 → 𝜓 ′ = 𝜓U † (46) an identical role in strong interaction processes,theproton and
The massive Dirac Lagrangian, the neutron are regarded as anisospin(or isobaric spin) doublet.
ℒ = −𝜓𝑖 𝜕𝜓𝑖 − 𝑚𝜓𝑖 𝜓𝑖 (47) 1
This doublet field,analogous to the 𝑠 = doublet of ordinary
1 2
This Lagrangian describesNspin- (anti)particlesof mass m
2 spin, is expressible as a column vector
and is invariant under local G transformations.If the ordinary 𝜓𝑝
derivative in (47) is replaced bya covariant derivative 𝜓= (54)
𝜓𝑛
(henceforth suppressing the indices i, j etc.) The transformation from one isospin state to other is
ℒ = −𝜓𝐷𝜓 − 𝑚𝜓𝜓 represented through the transformation matrix U which is
= −𝜓𝜕𝜓 − 𝑚𝜓𝜓 + 𝑔𝜓𝛾 𝜇 𝑊𝜇 𝜓, (48) related to transformation 𝜉 and SU(2) group generators ta
Therefore the last term denotes interaction between𝜓 and through Eq.(50). The transformation of the state 𝜓 to 𝜓 ′ can be
the gauge field𝑊𝜇 (𝑖. 𝑒. Wμ𝑎 t 𝑎 ). This term is also known as the represented as
interaction Lagrangian, can berewritten as 𝜓 → 𝜓′ = 𝑈𝜓, (55)
ℒ𝑖𝑛𝑡 = 𝑔Wμ𝑎 𝜓𝛾 𝜇 t 𝑎 𝜓 (49) It is a matter of convention which component of
The gauge transformations are unitary and t 𝑎 are the anti- ψcorresponds to the proton and which one tothe neutron
Hermitean generator matrices which are the parameters of the because of the almost exact symmetry of isospin invariance.
representation gauge group G. The matrices (ta)ijarethe non- The covariant derivative for transformation Eq.(55) is
abelian charges and are proportional to abelian charges straight forward, using Eq. (36), (51), (52) and (54), we get
𝑖𝑗 𝜕𝜇 𝜓𝑝 1 𝜓𝑝
𝐹𝑎 0 up to a proportionality factor i. The commutationrelation 𝐷𝜇 𝜓 = − 𝑖𝑔𝑊𝜇𝑎 𝜏𝑎 .
𝜕𝜇 𝜓𝑛 2 𝜓𝑛
for these charges is Lie algebra relation Eq.(5) which is
necessary and sufficient condition required for the non-abelian 𝜕𝜇 𝜓𝑝 1 𝑊𝜇3 𝑊𝜇1 − 𝑖𝑊𝜇2 𝜓𝑝
= − 𝑖𝑔 1 2 .
gauge transformations to constitute a group. 𝜕𝜇 𝜓𝑛 2 𝑊𝜇 + 𝑖𝑊𝜇 −𝑊𝜇3 𝜓𝑛
Now let us construct a gaugeinvariant Lagrangian for 1 1
𝜕𝜇 𝜓𝑝 − 𝑖𝑔𝑊𝜇3 𝜓𝑝 − 𝑖𝑔 𝑊𝜇1 − 𝑖𝑊𝜇2 𝜓𝑛
fermions transforming as doublets under the group = 2 2
SU(2)which, as we know, consists of all 2×2 unitary matrices 1 1
𝜕𝜇 𝜓𝑛 + 𝑖𝑔𝑊𝜇3 𝜓𝑛 − 𝑖𝑔 𝑊𝜇1 + 𝑖𝑊𝜇2 𝜓𝑝 (56)
with unit determinant. These matrices are expressible as an 2 2
exponential
15
IJFRCSCE | September 2018, Available @ http://www.ijfrcsce.org
_______________________________________________________________________________________
International Journal on Future Revolution in Computer Science & Communication Engineering ISSN: 2454-4248
Volume: 4 Issue: 9 12 – 20
_______________________________________________________________________________________________
The ordinaryderivative is replaced by a covariant one in Using the equation for the covariant derivative i.e. 𝐷𝜇 ≡
1
the Lagrangian of a degenerate doubletof spin − fields. We ∂𝜇 − gW𝜇 andalsothe equation for decomposition of the gauge
2
have, using Eq. (48), (49) and (51), field i.e.W𝜇 = W𝜇𝑎 𝑡𝑎 , we can write
ℒ = −𝜓𝐷𝜓 − 𝑚𝜓𝜓 𝐷𝜇 𝜑 ≡ ∂𝜇 𝜑 − gW𝜇𝑎 𝑡𝑎 𝜑
1 (57) 𝐷𝜇 𝜑 ∗ ≡ ∂𝜇 𝜑 ∗ − g𝜑 ∗ 𝑡𝑎† W𝜇𝑎
= −𝜓𝜕𝜓 − 𝑚𝜓𝜓 + 𝑖𝑔𝜓 W𝜇𝑎 𝛾𝜇 𝜏𝑎 𝜓, (63)
2
This forms what is known as a locally SU(2) invariant The transformation matrices U in (61) are unitary (so
Lagrangian.Note: Dirac 𝛾 𝜇 matrices are as below that 𝑡𝑎† = −𝑡𝑎 ). Therefore it can be shown using the inner
1 0 0 0 product 𝜙 2 = 𝜙𝑖∗ 𝜙𝑖 that the following Lagrangian is gauge
0 1 0 0 𝐼 0
𝛾0 = = invariant,
0 0 −1 0 0 −𝐼 2
0 0 0 −1 ℒ = − 𝐷𝜇 𝜙 − 𝑚2 𝜙 2 − 𝜆 𝜙 4 (64)
0 0 0 1 Substitution of Eq.(63) in Eq.(64), we get
1 0 0 1 0 = 0 𝜎1 ℒ = − ∂𝜇 𝜙 ∗ + g𝜑 ∗ 𝑡𝑎 W 𝑎𝜇 ∂𝜇 𝜙 − gW𝜇𝑏 𝑡𝑏 𝜙 − 𝑚2 𝜙 2
𝛾 =
0 −1 0 0 −𝜎1 0 −𝜆 𝜙 4
−1 0 0 0 2
ℒ = − ∂𝜇 𝜙 − 𝑚2 𝜙 2 − 𝜆 𝜙 4
− gW 𝑎𝜇 𝜙 ∗ 𝑡𝑎 ∂𝜇 𝜙 − ∂𝜇 𝜙 ∗ 𝑡𝑎 𝜙
0 0 0 −𝑖 + g 2 W 𝑎𝜇 W𝜇𝑏 𝜙 ∗ 𝑡𝑎 𝑡𝑏 𝜙
0 0 𝑖 0 0 𝜎2
𝛾2 = = This showsthe generators taare matrix
0 𝑖 0 0 −𝜎2 0
−𝑖 0 0 0 generalizations of the charge. It can be easily shown using
0 0 1 0 1
0 𝜎3 𝑡𝑎 = 𝑖𝜏𝑎 and𝜏𝑎 𝜏𝑏 + 𝜏𝑏 𝜏𝑎 = 2 𝐼 𝛿𝑎𝑏 that Lagrangian is
𝛾3 = 0 0 0 −1 = 2
−1 0 0 0 −𝜎3 0 (58) invariantunderSU(2).
0 1 0 0
2 1
Rewriting Eq.(39), we have ℒ = − ∂𝜇 𝜙 − 𝑚2 𝜙 2 − 𝜆 𝜙 4 − igW 𝑎𝜇 𝜙 ∗ 𝜏𝑎 ∂𝜇 𝜙
2 (65)
𝑎
𝐺𝜇𝜈 = ∂μ Wνa − ∂ν Wμa − 𝑔𝑓𝑏𝑐𝑎 Wμb Wνc 1 2 𝑎 2
− g W𝜇 𝜙2
Comparing Eq.(6) i.e. 𝑡𝑎 , 𝑡𝑏 = 𝑓𝑎𝑏 𝑐
𝑡𝑐 with Eq.(53), we 4
𝑐
get 𝑓𝑎𝑏 = −𝜀𝑎𝑏𝑐 , Therefore for SU(2) transformations, the field VI. THE GAUGE FIELD LAGRANGIAN
strength tensor is given by
𝑎 The locally invariant Lagrangians constructed so far in the
𝐺𝜇𝜈 = ∂μ Wνa − ∂ν Wμa + 𝑔𝜀𝑎𝑏𝑐 Wμb Wνc (59) preceding section arefor matter fields. The procedure is same
Using Eq.(20), we can show under infinitesimal SU(2) that the covariant derivative replaces theordinary one. The
transformations the gauge field Wμa transformsas gauge fields are considered responsible for interaction between
′
Wμa → Wμa = Wμa + 𝑔𝜀𝑎𝑏𝑐 Wμb ξc + 𝜕𝜇 ξ𝑎 (60) the matter fields for example fermions, quarks etc. They have
Weak interactions are governed by theories based on not yet been treatedas new dynamical degrees of freedom. They
SU(2) gauge transformations. Butthe strong forces are themselves are the fields and so can produce action and a force.
governed by SU(3) transformations.Since we use color as the That means they also have a Lagrangian, and therefore an
degrees of freedom transforming under SU(3) (like isospin in equation of motion. This Lagrangian should also be locally
SU(2)), theSU(3) gauge theory is known as quantum gaugeinvariant. This construction requires the use of the Lie
chromodynamicsor QCD in short. The corresponding gauge algebra valued field strength tensor Gμν. We have from Eq.(26)
′
fields in QCD are the gluon fields. The gluons are supposed to 𝐺𝜇𝜈 → 𝐺𝜇𝜈 = 𝑈𝐺𝜇𝜈 𝑈 −1
bind the quarks in hadrons.The quarks, like SU(2) doublet of The product of these tensors can be written as
proton and neutron (Eq.(54)) transform as triplets in SU(3) ′
Gμν Gρσ ⋯ Gλτ → Gμν ′
Gρσ ′
⋯ Gλτ = U Gμν Gρσ ⋯ Gλτ U−1 (66)
theory. Consequently the trace of these arbitrary products is
gauge invariant, i.e.
V. LAGRANGIAN FOR SPIN -0 FIELDS Tr Gμν Gρσ ⋯ Gλτ → Tr UGμν Gρσ ⋯ Gλτ U−1 = Tr Gμν Gρσ ⋯ Gλτ (67)
The similar procedure is adopted for determining Gauge by virtue of the cyclicity of the trace operation.
invariant Lagrangiansfor spin-0 fields.An array of complex Therefore 𝑇𝑟 𝐺𝜇𝜈 𝐺𝜇𝜈 is the simplest Lorentz invariant
scalar fieldsφ can be considered as a column vector. Let it and parity conserving term and
transforms under transformationsU. 1
𝜑 → 𝜑′ = 𝑈𝜑 (61) ℒ𝐺 = 2 𝑇𝑟 𝐺𝜇𝜈 𝐺𝜇𝜈 (67)
4𝑔
Then the complex conjugate field 𝜑 ∗ is a row vector, is the simplest possible Lagrangian for Gauge fields. The
𝜑 ∗ → (𝜑 ∗ )′ = 𝜑 ∗ 𝑈 † (62) 1
factor 2 is used for normalisation of the Lagrangian.The
4𝑔
other two simpler alternative Lorentz invariant
16
IJFRCSCE | September 2018, Available @ http://www.ijfrcsce.org
_______________________________________________________________________________________
International Journal on Future Revolution in Computer Science & Communication Engineering ISSN: 2454-4248
Volume: 4 Issue: 9 12 – 20
_______________________________________________________________________________________________
𝜇
forms 𝐺𝜇 and Gμν Gρσ εμνρσ ,are not used in the construction of complex conjugate matrices 𝑡𝑎∗ also generate a group
𝑅
the lagrangian. This is because the first one vanishesdue to the representation. Therefore, if𝑔𝑎𝑏 is complex, both its real and
anti-symmetry ofGμν. The form Gμν Gρσ εμνρσ does not conserve imaginary parts will separately constitute an invariant tensor.
𝑅
parity. It can be shown that it is actually equal to the total The invariant tensors 𝑔𝑎𝑏 depend on the representation,
divergenceof Chern–Simons term 𝜀 𝜇𝜈𝜌𝜎 𝑇𝑟 𝑊𝜈 𝜕𝜌 𝑊𝜎 − asfor an abelian group, Cartan-Killing metric (71) vanishes
2
trivially as the structure constants are zero. The abelian groups
𝑔𝑊𝜈 𝑊𝜌 𝑊𝜎 i.e. have nontrivial representations for which (70) does not vanish.
3
𝑅
𝜀 𝜇𝜈𝜌𝜎 𝑇𝑟 𝐺𝜇𝜈 𝐺𝜌𝜎 = 4 𝜕𝜇 𝜀 𝜇𝜈𝜌𝜎 𝑇𝑟 𝑊𝜈 𝜕𝜌 𝑊𝜎 The invariant tensors 𝑔𝑎𝑏 can be shown equal to proportionality
constant for the simple groups.
2 𝑅
− 𝑔𝑊𝜈 𝑊𝜌 𝑊𝜎 (68) The tensor 𝑔𝑎𝑏 or at least its real or imaginary part for any
3
and so it gets ignored while evaluating the action, at least in arbitrary representation can be diagonalisedwith eigenvalues
perturbation theory but not when one is dealing with non-trivial equal to 1,−1 or 0. This is achieved by suitably redefining the
a
boundary conditions. gauge fields and therefore, the group parametersξ and the
𝑅
Using Equation (39) and rescaling 𝑊𝜇 𝑡𝑜 𝑔𝑊𝜇 , we get the corresponding structure constants.This is because𝑔𝑎𝑏 has only
Lagrangian takes the form nonzero eigenvalues of equal sign that the first term in
𝐿𝑊 = 𝑇𝑟 (𝑡𝑎 𝑡𝑏 ) theLagrangian (69) is kinetic term for each of the gauge fields
𝑅
1 of thesame normalization. The tensor 𝑔𝑎𝑏 has non-positive
𝜕𝜇𝑊𝜈𝑎 − 𝜕𝜈 𝑊𝜇𝑎 𝜕𝜇 𝑊𝜈𝑏 − 𝜕𝜈 𝑊𝜇𝑏 eigenvalues. This is because for the so-called compact Lie
4
− 𝑔𝑓𝑐𝑑𝑎
𝑊𝜇𝑐 𝑊𝜈𝑑 𝜕𝜇 𝑊𝜈𝑏 groups with property𝑡𝑎† = −𝑡𝑎 .Further 𝑇𝑟 𝑡𝑎 𝑡𝑏 is replaced by
1 𝑎 𝑏 𝑓 −𝛿𝑎𝑏 in the Lagrangianas the lie groups are compact.Therefore
+ 𝑔2 𝑓𝑐𝑑 𝑓𝑒𝑓 𝑊𝜇𝑐 𝑊𝜇𝑒 𝑊𝜈𝑑 𝑊𝜈 (69) from Eq.(69) we get,
4
Therefore the Gauge field Lagrangian consists of a kinetic 1 2
𝐿𝑊 = − 𝜕𝜇 𝑊𝜈𝑎 − 𝜕𝜈 𝑊𝜇𝑎 + 𝑔𝑓𝑎𝑏𝑐 𝑊 𝑎𝜇 𝑊 𝑏𝜈 𝜕𝜇 𝑊𝜈𝑐
term, which is similar to one in the abelian Lagrangian, 4
a 𝑊 3 -interaction terms and a 𝑊 4 -interaction terms which 1
− 𝑔2 𝑓𝑎𝑏𝑐 𝑓𝑎𝑑𝑒 𝑊𝑏𝜇 𝑊𝜇𝑑 𝑊 𝑐𝜈 𝑊𝜈𝑒 (76)
depend on structure constants of the gauge group. 4
There is a tensor that multiplies the terms in The gauge-field Lagrangians for non-compact groups, for
aboveequation given as example for Lorentz group in the case of theory of Gravity,is
𝑅
𝑅
𝑔𝑎𝑏 = 𝑇𝑟 (𝑡𝑎 𝑡𝑏 ) (70) evaluated by evaluatingthe Feynman diagrams for 𝑔𝑎𝑏 having
Superscript R indicates that it depends on the non- zero eigen-values.However thegauge fields in the theory
of Gravity have rather different roles.
representation taken for the generators ta. The generators in the
Adding the gauge-field Lagrangian (76) with a gauge
adjoint representation definewhat is known asCartan-Killing
invariant Lagrangian for the matter fields, for example for
metric,
𝑐 𝑑
fermions in Eq.(48), we can write
𝑔𝑎𝑏 = 𝑓𝑎𝑑 𝑓𝑏𝑐 (71) ℒ = ℒ𝑊 + ℒ𝜓
These two representation equations define a totally anti- 1
symmetric tensor in all indices = 𝑇𝑟 (𝐺𝜇𝜈 𝐺𝜇𝜈 ) − 𝜓𝐷𝜓 − 𝑚𝜓𝜓 (77)
𝑅
𝑓𝑎𝑏𝑐 ≡ −𝑓𝑎𝑏 𝑑 𝑅
𝑔dc (72) 4
The field equations are the corresponding Euler-
This is because using (70), we can write Lagrangeequations evaluated using Hamilton’s principle These
𝑅 𝑑
𝑓𝑎𝑏𝑐 = −𝑓𝑎𝑏 𝑇𝑟 (𝑡𝑑 𝑡𝑐 ) = −𝑇𝑟 (𝑡𝑎 𝑡𝑏 𝑡𝑐 − 𝑡𝑏 𝑡𝑎 𝑡𝑐 ) (73) equations for fermions are covariant version of the Dirac
The right hand side of this equation is anti-symmetric equation, i.e.,
under the exchange of any two indices by virtue of cyclicity of 𝐷 + 𝑚 𝜓 = 0, 𝜓(𝐷 − 𝑚) = 0, (78)
the trace. †
𝑅 or using 𝑡𝑎 = −𝑡𝑎 , we can write
It is easy to show that, 𝑔𝑎𝑏 constitutes a
(𝜕 + 𝑚)𝜓 = 𝑔𝑊 𝑎 𝑡𝑎 𝛾 𝜇 𝜓 ,
groupinvarianttensorfor any representation. Arank-2 tensor
𝑇𝑎𝑏 inanadjoint representation𝑇𝑎𝑏 transformsas in Eq. (30) 𝜓 𝜕 + 𝑚 = −𝑔𝜓𝛾 𝜇 𝑡𝑎 𝑊𝜇𝑎 , (79)
𝑇𝑎𝑏 → 𝑇𝑎𝑏 − 𝜉 𝑐 (𝑓𝑐𝑎𝑑 𝑇𝑑𝑏 + 𝑓𝑐𝑏𝑑 𝑇𝑎𝑑 ) + 𝑂(𝜉 2 ) (74) These equations are the non-abelian generalizations of
Here as𝑔𝑎𝑏 𝑅
is a symmetric tensor as 𝑇𝑟 𝑡𝑎 𝑡𝑏 = 𝑇𝑟 𝑡𝑏 𝑡𝑎 . Let us abelian field equations.
substitute𝑔𝑎𝑏 𝑅
for𝑇𝑎𝑏 , we obtain The derivation of field equations for the gauge fields is
𝑅
𝑔𝑎𝑏 → 𝑔𝑎𝑏 𝑅 𝑅
− 𝜉 𝑐 (𝑓𝑐𝑎𝑏 + 𝑓𝑐𝑏𝑎𝑅
) + 𝑂(𝜉 2 ) more complicated. Let us consider an infinitesimal
𝑅 𝑅 𝑅 transformation causing a small variation in the field 𝑊𝜇
Or 𝛿𝑔𝑎𝑏 → −𝜉 (𝑓𝑐𝑎𝑏 + 𝑓𝑐𝑏𝑎 ) + 𝑂(𝜉 2 )
𝑐 (75)
But 𝛿𝑔𝑎𝑏 𝑅
vanishes because of the total anti-symmetry 𝑊𝜇 → 𝑊𝜇 + 𝛿𝑊𝜇 (80)
𝑅
of 𝑓𝑎𝑏𝑐 . This proves the invariance of the tensors 𝑔𝑎𝑏 𝑅
. Here 𝛿𝑊𝜇 is a covariant quantity transforming in the
∗
The complex conjugate matrices 𝑡𝑎 also satisfy the Lie- adjointrepresentation.
algebra Eq.(5) with real structure constants. Therefore the The corresponding change in field strength is

17
IJFRCSCE | September 2018, Available @ http://www.ijfrcsce.org
_______________________________________________________________________________________
International Journal on Future Revolution in Computer Science & Communication Engineering ISSN: 2454-4248
Volume: 4 Issue: 9 12 – 20
_______________________________________________________________________________________________
1b a cμν
𝐺𝜇𝜈 → 𝐺𝜇𝜈 + 𝐷𝜇 (𝛿𝑊𝜈 ) − 𝐷𝜈 (𝛿𝑊𝜇 ), (81) = − gGμν fbc G
2
Where covariant derivative on 𝛿𝑊𝜈 is defined as Right hand side vanishes because of anti-symmetricnature of
𝐷𝜇 (𝛿𝑊𝜈 ) = 𝜕𝜇 (𝛿𝑊𝜈 ) − 𝑔 [𝑊𝜇 , 𝛿𝑊𝜈 ] (82) 𝑓𝑏𝑐𝑎 .Therefore from Eq.(93), we get
AsGμνis anti-symmetric in μ and ν, the variation in action due 𝐷𝜇 𝐽𝜇𝑎 = 0 (95)
to this transformation can be shown as Or raising indices by two and rescaling W to gW, we have
1 from Eq.(34) that
𝛿 𝑑4 𝑥𝑇𝑟 𝐺𝜇𝜈 𝐺𝜇𝜈 = 𝑑4 𝑥𝑇𝑟 𝐺𝜇𝜈 𝐷𝜇 𝛿𝑊𝜈 (83) 𝐷𝜇 = ∂μ − 𝑔𝑓𝑏𝑐𝑎 𝑊𝜇𝑏
4
The integrand on the right-hand side can be written as Therefore from Eq.(95), we get
𝐷𝜇 𝑇𝑟 𝐺𝜇𝜈 𝛿𝑊𝜈 − 𝑇𝑟 𝐷𝜇 𝐺𝜇𝜈 𝛿𝑊𝜈 (84) 𝜕𝜇 𝐽𝜇𝑎 − 𝑔𝑓𝑏𝑐𝑎 𝑊𝑏𝜇 𝐽𝜇𝑐 = 0 (96)
𝑇𝑟 𝐺 𝛿𝑊𝜈 is gauge invariant as δWμis a covariant object.
𝜇𝜈 The above result in Eq. (95) shows that the gaugefields
couple tocurrents thatare covariantly constant. However the
Therefore replacing covariant derivative by an ordinary
charges associated with the current are not quite conserved as
derivative and using Gauss’ theorem, the first term converts
evident from Eq.(96). This is obviously contribution of non-
into a surface integral over the boundary of the integration
neutral gauge fields must be included to define charges for their
domain in (83). The variation δWμhowever,vanishes at the conservation.
boundariesin accordance with Hamilton’s principle. Therefore, VII. CHIRAL GAUGE GROUPS AND DISCRETE
only the second term in (84) contributes. Therefore from
SYMMETRIES
Eq.(83) , we have
There is lot of literature available on topics such as
𝛿 𝑑4 𝑥ℒ𝑤 = − 𝑑4 𝑥𝑇𝑟 𝐷𝜇 𝐺𝜇𝜈 𝛿𝑊 𝜈 (85) discrete symmetries, like parity reversal and charge
Similarly the variation of the matter action due to (80) is conjugation, also on topics like chiral spinorfields,
given as Eigenspinors of the γ5 matrix. We skip our discussion on these
terms.
𝛿 𝑑4 𝑥ℒψ = − 𝑑4 𝑥𝑇𝑟 𝛿𝑊𝜇 𝐽𝜇 (86)
Let us consider non-abelian group of gauge
Changing dummy index from 𝜇 to 𝜈and adding Eq.(86) toEq.(85), we transformations based on generators tathat give rise to
get fermionic vector or axial-vector currents. The axial vectors are
𝛿 𝑑4 𝑥 ℒ𝑤 + ℒψ = − 𝑑4 𝑥𝑇𝑟 𝐷𝜇 𝐺𝜇𝜈 + 𝐽𝜈 𝛿𝑊 𝜈 those quantities which do not change direction under mirror
The variation in Lagrangian should be zero, therefore the field reflection such as angular velocity of rotating wheel of a
equation for the gauge field are receding car on road. More on axial vectors can be seen in the
𝐷𝜇 𝐺𝜇𝜈 + 𝐽𝜈 = 0 , (87) literature available on internet.
Both 𝐺𝜇𝜈 and 𝐽𝜇 are Lie-algebra valued matrices. Using Further the matrices𝑡𝑎 𝛾5 act on both the gauge-group and
the spinor indicesbutdo not generate a group of
𝑇𝑟 𝑡𝑎 𝑡𝑏 = −𝛿𝑎𝑏 and 𝐽𝜇 = 𝐽𝜇𝑎 𝑡𝑎 the variation of the gauge field
transformations.This is because the group of matrices 𝑡𝑎 𝛾5 is
and the matter action can be written as
not closed under commutation
𝛿 𝑑4 𝑥ℒ𝑤 = 𝑑4 𝑥𝛿𝑊 𝑎𝜈 𝐷𝜇 𝐺𝜇𝜈
𝑎
(88) [𝑡𝑎 𝛾5 , 𝑡𝑏 𝛾5 ] = 𝑓𝑎𝑏 𝑐 𝑐
𝑡𝑐 ≠ 𝑓𝑎𝑏 𝑡𝑐 𝛾5 (97)
The dimension of the group and thus the number of
𝛿 𝑑4 𝑥ℒψ = 𝑑4 𝑥𝛿𝑊𝜇𝑎 𝐽𝜇𝑎 (89)
independent gauge field doubles if both taandtaγ5are taken as
Therefore, in terms of components, we get independent generators.Therefore chiral projection operators
Dν Gμν
a
= J𝜇a (90) defined by P± = (1 ± γ5 )
1
are used. These
2
Writing in terms of Gauge field𝑊, we get a rather complicated 2
Projectionoperators satisfy𝑃± = 𝑃± , and therefore the matrices
equation
𝜕 𝜈 𝜕𝜇𝑊𝜈𝑎 − 𝜕𝜈 𝑊𝜇𝑎 + 𝑔𝑓𝑏𝑐 𝑎
𝑊 𝑏𝜈 𝜕𝜇 𝑊𝜈𝑐 + 𝑊𝜇𝑏 𝜕𝜈 𝑊 𝑐𝜈
𝑡𝑎 𝑃± do generate a representation group. We can verify the
− 2𝑊 𝜕𝜈 𝑊𝜇𝑐 + 𝑔2 𝑓𝑏𝑐
𝑏𝜈 𝑎 𝑐
𝑓𝑑𝑒 𝑊𝜇𝑑 𝑊𝜈𝑏 𝑊 𝑒𝜈 = J𝜇a
(91) matrices 𝑡𝑎 𝑃± satisfy lie-algebra i.e.
𝑐
(𝑡𝑎 𝑃± ), (𝑡𝑏 𝑃± ) = 𝑓𝑎𝑏 𝑡𝑐 𝑃± (98)
The current J𝜇a can be evaluated as
The fermion fieldsdecompose into chiral components ψL=
J𝜇a = 𝑔𝜓𝛾𝜇 𝑡𝑎 𝜓 (92) P+ψand ψR= P−ψ. These chiral components ψLandψRform
Applying a covariant derivative Dμand raising indices by independent representations of the gauge group.Also 𝑡𝑎R and𝑡𝑎L ,
two in Eq.(90), we get which are a priori unrelated but satisfythe same commutation
𝜇 𝑎
Dμ Dν G aμν = 𝐷 𝐽𝜇 (93) relations as given by Eq.(98).These two sets of generators of
On left hand side using antisymmetry of Gaμνand Ricci the gauge group act on independent arrays of fields denoted by
identity (27), we get ψRand ψL, respectively. Thecoupling of the gauge fields to the
fermions thereforeis represented in terms of the
1
1 chiralprojectors P± = (1 ± γ5 ) and is proportional to
Dμ Dν G aμν = D , D G aμν (94) 2
2 μ ν 𝛾𝜇 1 + 𝛾5 𝑡𝑎𝐿 + 𝛾𝜇 (1 + 𝛾5 )𝑡𝑎𝑅 .
18
IJFRCSCE | September 2018, Available @ http://www.ijfrcsce.org
_______________________________________________________________________________________
International Journal on Future Revolution in Computer Science & Communication Engineering
Volume: 4 Issue: 9
_______________________________________________________________________________________________
There are a number of possibilities.Firstly𝑡𝑎R and 𝑡𝑎L arein-
equivalent. Then the numbers of left- andright-handed fields is
not equal. Second possibility is that only one of 𝑡𝑎R or 𝑡𝑎L is
present, then the gauge fields coupleto a vector or an axial
vector, i.e. eitherto𝛾𝜇 1 + 𝛾5 or to 𝛾𝜇 1 − 𝛾5 . That means the
only field present is of one chirality. The third possibility is that
ψLandψRtransformequivalently, or 𝑡𝑎L = 𝑡𝑎R . This always leads
to a purely vector-like coupling. The fourth possibility is that
the generators are each other’s complexconjugates 𝑡𝑎L =
𝑡𝑎R ∗. Acomplex conjugate representation is always possible
as the complexconjugate generators satisfy the same
commutator algebra
𝑡𝑎 ∗ , 𝑡𝑏 ∗ = 𝑓𝑎𝑏𝑐
𝑡𝑐 ∗ (99)
The representation is real so it is possible that all
generators are imaginary. An argument similar to Eq.(97)
shows that this is possible only if the group is abelian.Therefore
the gauge fieldsassociated with real generators have vectorlike
and those with imaginary generators haveaxial-vectorlike
coupling to fermions.
The generators for charge-conjugated fields 𝜓𝐿𝑐 and 𝜓𝑅𝑐 ,
are the complex conjugate matrices
𝑡𝑎R ∗ and 𝑡𝑎L ∗ respectively. Thereforeif the left-handed
fields, 𝜓𝐿 and 𝜓𝐿𝑐 form a basisof independent fermion fields, the
reducible representation with generators is given as
𝑡L𝑎 0 right-handed representations. In the first case the gauge
∗ (100)
0 𝑡R𝑎
These matrices also satisfy therelevant commutation
relations associated with the gauge group. Similar
representation exists for the right-handed fieldsψRand ψcR.
However, the gauge transformations act identically on right-
andleft-handed components and the representation is called as
vector-like. This is because thecorresponding fermionic
currents contain no axial vector like term 𝛾𝜇 𝛾5 terms because
the generators in Eq.(100) can bewritten in real form by an
appropriate change of basis. Therefore the representationbased
on these generators is real. Further it is possible to write a
standard gauge-invariant mass term for a real representation
which involves the products of right- and left-handed spinors.
Thevectors like representationsarealso free of so-called
anomalies.
The invariance of gauge theory under
discretetransformations such as P, C or the combined
transformation CP, requires that the covariantderivatives of the
spinors transform just as ordinary derivatives underthe
transformations. The covariant derivatives of 𝜓𝐿 and
𝜓𝑅 containmatrices W𝜇L = 𝑊𝜇𝑎 𝑡𝑎𝐿 and W𝜇R = 𝑊𝜇𝑎 𝑡𝑎𝑅 and read
Dμ ψL = ∂μ ψL − WμL ψL , Dμ ψR = ∂μ ψR − WμR ψR (101)
Therefore transformationrules for the gauge fields can be
determined.
Let us start with parity reversal, under which left- and
righthandedspinors are interchanged.The gauge fields under
parity reversal must transform as,
IJFRCSCE | September 2018, Available @ http://www.ijfrcsce.org
_______________________________________________________________________________________
ISSN: 2454-4248
12 – 20
𝑃
𝑊𝜇𝑎 (𝒙, 𝑥 0 ) 𝑡𝑎𝐿 → −(1 − 2𝛿𝜇 0 )𝑊𝜇𝑎 (−𝒙, 𝑥 0 ) 𝑡𝑎𝑅 (102)
and vice versa for the covariant derivatives of ψLandψRto
transform into each other. The action of P on the gauge fields is
fixed, and adding an extra phase factor to it is not possibleif the
gauge-group representation is once chosen. However, Eq. (102)
requires that the generators 𝑡𝑎𝐿 and 𝑡𝑎𝑅 are linearly dependent.
Therefore there are twosituations.The representations are
equivalent or each other’s complex conjugate. The gauge
fieldstransform as vectors under parity reversal and the theory
is vectorlike in the first case. The gauge fields, in the second
case, associated with the real generators transform asvectors
while gauge fields associated with imaginary generators
transform asaxial-vectors underparity reversal.
Thecovariant derivatives under charge conjugation, are
given by the expression as below
𝐷𝜇 𝜓𝐿𝑇 = 𝜕𝜇 𝜓𝐿𝑇 − 𝑊𝜇𝐿 𝜓𝐿𝑇
Or
𝜓𝐿 𝐷𝜇 = 𝜓𝐿 𝜕𝜇 − 𝑊𝜇†
The gauge field must transform under charge
conjugationas below to obtain the correct transformation of the
covariant derivative
𝐶
𝑊𝜇𝑎 𝑥, 𝑥 0 𝑡𝑎𝐿 → 𝑊𝜇𝑎 𝑥, 𝑥 0 𝑡𝑎𝑅 ∗ (103)
This result is meaningful for (i) equivalent left- and right-
handedrepresentations and (ii)complex conjugate left- and
fieldsassociated with real generators are even and gauge fields
with purely imaginarygenerators are odd under charge
conjugation. In the second case all fieldsare even.
Theparityand charge conjugation can be separately
defined if spinors of bothchirality are present.However this is
not so for CP conjugation. Combining the previous results Eq.
(102) and Eq.(103) theCPtransformation of the gauge fields is
given by
𝐶𝑃
𝑊𝜇𝑎 𝑥, 𝑥 0 𝑡𝑎 −(1 − 2δμ0 )𝑊𝜇𝑎 −𝑥, 𝑥 0 𝑡𝑎 ∗ (104)
Again there is no possibility for assigning an arbitrary phase
factor. The invariance under CP requires the relation (104) to
hold good for all irreducible representations.The CP
phasefactor clearly depends on whether the gauge field has a
real or purelyimaginary generator.
CONCLUSIONS
The non-abelian transformations produce different
outcomes if the order of transformations is changed. The
generators of the representation group of non-abelian
𝑐
transformations follow Lie algebra 𝑡𝑎 , 𝑡𝑏 = 𝑓𝑎𝑏 𝑡𝑐 and are
therefore non-commuting.Non-abelian transformations in
isospin space SU(2) successfully generates Lagrangian for spin
-1/2, spin 0 and also for Gauge fields, which are considered
responsible for interaction between the matter fields. Non-
abelian transformations in SU(3) space known as
chromodynamics explain gluon interactions between the Quark
fields. Discrete symmetries like parity and charge conjugation
are also explained by the non-abelian transformations
19
International Journal on Future Revolution in Computer Science & Communication Engineering ISSN: 2454-4248
Volume: 4 Issue: 9 12 – 20
_______________________________________________________________________________________________
ACKNOWLEDGMENT
Author will like to thank Head, Department of Physics,
Kumaun University, Nainital, Uttarakhand and Principal, R.H.
Govt PG College, Kashipur (U.S. Nagar) for their generous
support in perusing these studies.

REFERENCES
[1] B. de Wit and J. Smith, Field theory in particle physics, Vol 1 and
2 (North—Holland, 1986).
[2] J .C. Taylor, Gauge theories of weak interactions (Cambridge
Univ. Press, 1976)
[3] C. Quigg, Gauge theories of the strong, weak and electromagnetic
interactions (Benjamin, 1983)
[4] F. Mandl and G. Shaw, Quantum field theory (Wiley, 1984)
[5] K. Huang, Quarks, leptons and gauge fields (World Scient., 1981)
[6] F. Halzen and A. Martin, Quarks and leptons (Oxford, 1984)
[7] I.J.R. Aitchison and A. Hey, Gauge theories in particle physics
(Hilger, 1982)

Dr. Dev Raj Mishra won Gold

Medal at Graduation for standing First
in the University, Did M.Sc. (Physics)
from M.D. University, Rohtak
(Haryana) and did his Ph.D. in High
Temperature Superconductivity from
University of Delhi, Delhi. He worked
as research fellow during his Ph.D. in
Cryogenics Group in National
Physical Laboratory and visited for about Eight months in
Low Temperature Laboratory at Helsinki University of
Technology, Espoo, Finland. Joined as Lecture in Department
of Higher Education, Government of Uttarakhand, INDIA in
Jan. 2000 and presently posted at Department of Physics,
R.H. Govt.P.G. College,Kashipur, US Nagar, Uttarakhand
(INDIA) -244713.

20
IJFRCSCE | September 2018, Available @ http://www.ijfrcsce.org
_______________________________________________________________________________________