Spontaneous breaking of continuous symmetriesdermisek/QFT_09/qft-II-6-4p.pdf · Spontaneous...
-
Upload
truongthuy -
Category
Documents
-
view
221 -
download
4
Transcript of Spontaneous breaking of continuous symmetriesdermisek/QFT_09/qft-II-6-4p.pdf · Spontaneous...
Spontaneous breaking of continuous symmetriesbased on S-32
Consider the theory of a complex scalar field:
the lagrangian is invariant under the U(1) transformation:
for negative, the minimum of the potential
is achieved for
we have a continuous family of minima (ground states):
327
is achieved for
we have a continuous family of minima (ground states):
all the minima are equivalent, there is a flat direction in the
field space (we can move along without changing the energy)
the physical consequence of a flat direction is the existence of a massless particle - Goldstone boson
for we can write:
b is massless!
and we get:
real scalar fields
328
It is also instructive to use a different parameterization:real scalar fields
now we get:
the U(1) transformation takes a simple form:
parameterizes the flat direction
is massless!spectrum and scattering amplitudes are independent of field redefinitions!
The Goldstone boson remains massless even after including loop corrections!
329
To see it, note that if the GB remains massless, the exact propagator:
has to have a pole at .
The vertices are proportional to momenta of (because of derivatives) and so they vanish.
we can calculate it by summing 1PI diagrams with two external lines with .
Note: the theory expanded around the minimum of the potential has interactions that were not present in the original theory. However it turns out that divergent parts of the counterterms for the theory with (where the symmetry is unbroken) will also serve to cancel the divergencies in the theory with where the symmetry is spontaneously broken. This is a general rule for theories with spontaneous symmetry breaking.
330
Let’s now generalize our results for a nonabelian case:
the lagrangian is invariant under the SO(N) transformation:
set of antisymmetric generator matrices with a single nonzero entry -i above the main diagonal
set of infinitesimal parameters
for , the minimum of the potential is achieved for:
with
the vacuum expectation value, N-component vector with arbitrary direction
we can choose the coordinate system (SO(N) rotation) so that
331
Let’s make an infinitesimal SO(N) transformation; this changes the VEV:
zero for all the generators with no non-zero entry in the last column - UNBROKEN GENERATORS
non-zero for N-1 generators with a non-zero entry in the last column - BROKEN GENERATORS
these do not change the VEV of the field, and correspond to the unbroken symmetry
these change the VEV of the field but not the energy; each of them corresponds to a flat direction in
field space; each flat direction implies the existence of a massless
particle - Goldstone boson.For SO(N) we have
unbroken generators SO(N-1) - obvious symmetry of the lagrangian!
332
Let’s work it out in detail; we can write our lagrangian as:
where
summed from 1 to N
Let’s shift the field by the VEV and define:
we get:
where: summed from 1 to N-1
SO(N-1) - manifest symmetry of the lagrangian!
N-1 massless particles - Goldstone bosons
333
Spontaneous breaking of gauge symmetriesbased on S-84,85
Consider scalar electrodynamics:
where
for , the minimum of the potential is achieved for:
we write:
and the scalar potential becomes:
334
The gauge symmetry allows us to shift the phase of by an arbitrary spacetime function; we can set:
with this choice the kinetic term becomes:Unitary gauge
mass term for the gauge field!
Higgs mechanism: the Goldstone boson disappears and the gauge field acquires a mass. The Goldstone boson has become the longitudinal polarization of the massive gauge field.
The scalar field that is used to break the gauge symmetry is called the Higgs field.
335
Part of the lagrangian quadratic in gauge fields:
the equation of motion:
acting on both sides with we find:
thus each component obeys the Klein-Gordon equation:
this is not a gauge-fixing condition!
the general solution is:
with
336
the general solution is:
in the rest frame we can choose the polarization vector to correspond to definite spin along the z-axis:
polarization vectors together with the timelike unit vector form an orthonormal and complete set:
337
In analogy with the scalar field theory or QED the propagator for the “massive” gauge field is:
and interactions can be read off of:
and
338
taking the unitary gauge, , corresponds to inserting a functional delta function:
There are complications with calculating loop diagrams!
to the path integral. Changing integration variables from and to and we must include the functional determinant:
arbitrary constant
ghost propagator:
ghost-ghost-scalar vertex:
no ghost-gauge field interactionsthe ghost propagator doesn’t contain momentum and so loop diagrams with arbitrarily many external lines diverge more and more (also the gauge field propagator scales like for large momenta); difficult to establish renormalizability.
339
The gauge doesn’t suffer from this problem:real scalar fields
with this parameterization, the potential is:
and the covariant derivative is:
and so the kinetic term can be written as:
340
mass term for b
rearranging the kinetic term:
cross term between the vector field and
the b-field
For abelian gauge theory in the absence of spontaneous symmetry breaking we fixed the gauge by adding the gauge-fixing and ghost terms:
ghost fields have no interactions and can be ignored.
For spontaneously broken theory we choose:
and we get:
canc
el
integration by parts
341
Let’s choose:
gauge-fixing function
we will get the gaugethen
-1 for acting as an anticommuting object
or REVIEW
342
now we can easily perform the path integral over B:
it is equivalent to solving the classical equation of motion,
and substituting the result back to the formula:
we obtained the gauge fixing lagrangian and the ghost lagrangian
REVIEW
343
and for the ghost term we find:
For spontaneously broken theory we choose:
ghost has the same mass as the b-field!
344
Collecting terms quadratic in the vector field:
in the momentum space we have:
which can be easily inverted:
propagator for the massive vector field in the gauge
345
Propagator highly simplifies for :
propagator for the massive vector field in the gauge:
transverse components propagate with mass M
longitudinal component propagate with mass(the same as ghost and b)
since scattering amplitudes do not depend on they all are unphysical particles
typically used for loop calculations
346
Summary of gauge:
interactions are collected here:
ghost and the b-field propagators:
physical Higgs field with mass:
gauge field (3 polarizations) with mass:
347
It is instructive to consider limit:
gauge field propagator:
this is what we found in the unitary gauge
ghost and the b-fields become infinitely heavy
but we cannot just drop the ghost field because its interaction becomes infinite as well; instead we can define:
and consider the ghost propagator to be and the vertex
this is what we found in the unitary gauge
The in the limit is equivalent to unitary gauge!
348
Spontaneously broken nonabelian gauge theorybased on S-84,86
Generalization to a nonabelian gauge theory is straightforward:
the potential is minimized for:
plugging the vev to the kinetic term we get a mass term for gauge fields:
where
349
gauge fields corresponding to unbroken generators, , remain massless
plugging the vev to the kinetic term we get a mass term for gauge fields:
where
gauge fields corresponding to broken generators, , get a mass
(and they form representations of the unbroken group)
(these are the gauge fields of unbroken gauge symmetry)
we can repeat everything that we did for the abelian case, see S-86, instead of working out details for the general case let’s focus on a specific example...
350
The Standard Model: gauge and Higgs sectorbased on S-87
The Standard Model of elementary particles describes almost everything we know about our world. It is based on SU(3)xSU(2)xU(1) gauge symmetry spontaneously broken to SU(3)xU(1).
All particles acquire mass from interactions with the Higgs field in the process of electroweak
symmetry breaking
Glashow-Weinberg-Salam model
351
the covariant derivative is:
can be conveniently written as
Let’s start with the electroweak part which is described by SU(2)xU(1) gauge symmetry and one complex scalar field - the Higgs field in the representation:
352
the potential is (the usual one):
the scalar field get a vev, and we can use a global SU(2) transformation to bring it to the form:
plugging the vev to the kinetic term:
we get the mass term for gauge fields:
353
we have to diagonalize the mass matrix for gauge fields, we define:
then we have
the weak mixing angle
remains massless, it is the gauge field of
354
let’s work out the complete lagrangian:we will use unitary gauge
(it is simple and sufficient for tree-level calculations)
a complex scalar field (which is a doublet under SU(2)) can be written in terms of 4 real scalar fields; three of them become the longitudinal components of W and Z, thus we can write the Higgs doublet as: real scalar field
the Higgs field has the usual kinetic term and the potential is now:
we obtained the mass terms for gauge fields by inserting just the vev; to get the whole lagrangian that contains also interactions of Higgs and gauge fields we simply multiply the mass term for gauge fields by
355
we can combine
finally, let’s look at the kinetic terms for the gauge fields:
where
where
and we get a part of the lagrangian:
356
we also have:
where:
and we get another part of the lagrangian:
357
Putting all the terms together:
charge of is ,e is positive (opposite sign that we used in QED)
The lagrangian exhibits manifest gauge invariance!
358
The Standard Model: Lepton sectorbased on S-88
as far as we know there are three generations of leptons:
and they are described by introducing left-handed Weyl fields:
just a name
359
the covariant derivatives are:
and the kinetic term is the usual one for Weyl fields:
there is no gauge group singlet contained in any of the products:
it is not possible to write any mass term!
360
Lagrangians for spinor fields
we want to find a suitable lagrangian for left- and right-handed spinor fields.
based on S-36
Lorentz invariant and hermitian
it should be:
quadratic in and equations of motion will be linear with plane wave solutions
(suitable for describing free particles)
terms with no derivative:
terms with derivatives:
+ h.c.
would lead to a hamiltonian unbounded from belowREV
IEW
361
to get a bounded hamiltonian the kinetic term has to contain both and , a candidate is:
is hermitian up to a total divergence
does not contribute to the action
are hermitianREVIEW
362
Taking hermitian conjugate:
Our complete lagrangian is:
the phase of m can be absorbed into the definition of fields
and so without loss of generality we can take m to be real and positive.
Equation of motion:
are hermitianREVIEW
363
however we can write a Yukawa term of the form:
there are no other terms that have mass dimension four or less!
in the unitary gauge we have:
and the Yukawa term becomes:
364
it is convenient to label the components of the lepton doublet as:
then we have:
and we see that the electron has acquired a mass:
and the neutrino remained massless!
we define a Dirac field for the electron
365
consider a theory of two left-handed spinor fields:
i = 1,2
the lagrangian is invariant under the SO(2) transformation:
it can be written in the form that is manifestly U(1) symmetric:
REVIEW
366
Equations of motion for this theory:
we can define a four-component Dirac field:
Dirac equation
we want to write the lagrangian in terms of the Dirac field:
Let’s define:numerically
but different spinor index structureREVIEW
367
Then we find:
Thus we have:REVIEW
368
Thus the lagrangian can be written as:
The U(1) symmetry is obvious:
The Nether current associated with this symmetry is:
later we will see that this is the electromagnetic currentREVIEW
369
If we want to go back from 4-component Dirac or Majorana fields to the two-component Weyl fields, it is useful to define a projection matrix:
just a name
We can define left and right projection matrices:
And for a Dirac field we find:
REVIEW
370
We can describe the neutrino with a Majorana field:
and it is also convenient to define:
because then the kinetic term:
can be written as:
a “Dirac field” for the neutrino
371
Now we have to figure out lepton-lepton-gauge boson interaction terms:
(we want to write covariant derivatives in terms of , and )
we have
and
we identify it with electric charge:
372
since we have:
the electric charge assignments are as expected:
covariant derivatives in terms of the four-component fields:
373
Putting pieces together,
we find the interaction lagrangian:
where
374
The connection with Fermi theory:
if we want to calculate some scattering amplitude (or a decay rate) of leptons with momenta well below masses of W and Z, the propagators effectively become:
and we get 4-fermion-interaction terms:
where the Fermi constant is:
375
Generalization to three lepton generations:
the most general Yukawa term:
a complex 3x3 matrix;it can be diagonalized (with positive or zero entries on the diagonal)by a bi-unitary transformation
fermion masses are then given by:
diagonal entries
currents remain diagonal, just add the generation index
376
We can calculate e.g. muon decay:
the relevant part of the charged current:
we get the effective lagrangian:
and the rest is straightforward...
377
The Standard Model: Quark sectorbased on S-89
as far as we know there are three generations of quarks:
and they are described by introducing left-handed Weyl fields:
just a name
378
the covariant derivatives are:
and the kinetic terms are the usual ones for Weyl fields:
there is no gauge group singlet contained in any of the products;
it is not possible to write any mass term!
379
however we can write a Yukawa terms of the form:
there are no other terms that have mass dimension four or less!
in the unitary gauge we have:
and the Yukawa terms become:
380
it is convenient to label the components of the quark doublet as:
then we have:
and we see that the up and down quarks have acquired masses:
...work out interactions with gauge fields, and generalization to three families...
we define Dirac fields for the down and up quarks
381