inelastic scattering

2004,Torino Aram Kotzinian 1

Inelastic scattering

When the scattering is not elastic (new particles are produced) the energy and direction of the scattered electron are independent variables, unlike the elastic scattering situation.W is the mass squared of the produced hadronic systemFrom the measurement of the direction (solid angle element d) and the energy E' of the scattered electron, the four momentum transfer Q2=-q2 can be calculated.The differential cross-section is determined as a function of E' and Q2.

q2 4 EE ' sin2 2

e (k,E)

N (P,M)

e(k',E')

(q)

W

k ' k q

W 2 (P q)2

P2 2Pq q2

M 2 2M Q2

'EE


Electron - proton inelastic scattering

Bloom et al. (SLAC-MIT group) in 1969 performed an experiment with high-energy electron beams (7-18 GeV).

Scattering of electrons from a hydrogen target at 60 and 100.

Only electrons are detected in the final state - inclusive approach.

The data showed peaks when the mass W of the produced hadronic system corresponded to the mass of the known resonances.


Inelastic scattering cross-section

The cross-section is double differential because and E ' are independent variables.The expression contains Mott cross-section as a factor and is analogous to the Rosenbluth formula. It isolates the unknown shape of the nucleon target in two structure functions W1 and W2, which are the functions of two independent variables and q2. The structure functions correspond to the two possible polarisation states of the virtual photon: longitudinal and transverse. Longitudinal polarisation exists only because photon is virtual and has a mass.For elastic scattering, (P+q)2=M2 and the two variables and Q2 are related by Q2=2M .

Similar to the electron-proton elastic scattering, the differential cross-sectionof electron-proton inelastic scattering can be written in a general form:

dddE '

2 cos2 2

4E 2 sin4 2

W2(,q2 ) W1(,q2 )tan2 2


Scaling

To determine W1 and W2 separately it is necessary to measure the differential cross-section at two values of and E' that correspond to the same values of and Q2.

This is possible by varying the incident energy E.

SLAC result: the ratio of /Mott depends only weakly on Q2 for high values of W.

For small scattering angles /Mott ≈ W2 . Thus, the structure function W2 does not depend on Q2.


Scaling

Instead, at high values of W the function W2 depends on the single variable = 2M / Q2 (at present the variable x=1/ is widely used)This is the so-called "scaling" behaviour of the cross-section (structure function).It was first proposed by Bjorken in 1967.

W1,2(,q2) W1,2(x) when ,q2 ∞.


Deep Inelastic Scattering (DIS)Kinematic Variables

M --The mass of the target hadron. E -- The energy of the incident lepton.k -- The momentum of the initial lepton. -- The solid angle into which the outgoing lepton is scattered.E’ -- The energy of the scattered lepton.K’ -- The momentum of the scattered lepton,

K’ = (E’;E’sincos;E’sinsin;E’cos).P -- The momentum of the target, p = (M; 0; 0; 0), for a fixed target experiment.q = k-k’ -- the momentum transfer in the scattering process, i.e. the momentum of the virtual photon.

z-axis to be along the incident lepton beam direction.


Important variables

The invariant mass of the final hadronic system X is


Some inequalitiesThe invariant mass of X must be at least that of a nucleon, since baryon number is conserved in the scattering process.

Since Q2 and n are both positive, x must also be positive. The lepton energy loss E-E’ must be between zero and E, so the physically

allowed kinematic region is

The value x = 1 corresponds to elastic scattering.


Any fixed hadron state X with invariant mass contributes to the cross-section at the value of x

In the DIS limit

So, any hadron state X with fixed invariant mass gets driven to x=1

The experimental measurements give the cross-section as a function of the final lepton energy and scattering angle. The results are often presented instead by giving the differential cross-section as a function of (x, ) or (x,y). The Jacobian for converting between these cases is easily worked out using the definitions of the kinematic variables

XM

2Q

xQ fixed with ,2


Thus the cross-sections are related by

the contours of constant x are straight lines through the origin with slope x.

EMEQ

MQ

x

22

22

)cos1)((1

2

2

EMME

Q

the contours of constant angle are straight lines passing through the point


It is useful to have formulae for the different components of q as a function of x and y.

For fixed value of x, the maximum allowed value of

2Q

This expressions are valid in the Lab frame with z-axis along lepton momentum


Expression for DIS cross sectionThe scattering amplitude MM is given by

onpolarizati target -

onpolarizatilepton -

ls


It is conventional to define the leptonic tensor

The definition of the hadronic tensor is slightly more complicated.

Inserting a complete set of states gives

where the sum on X is a sum over the allowed phase space for the final state X


Only the first term contributes, since and Using leptonic and hadronic tensors we have

Translation invariance implies that

00 ppX 00 q


Finally, integrating over azimuth, we get


Leptonic tensor

The polarization of a spin 1/2 particle can be described by a spin vector defined in the rest frame of the particle by

The spin vector in arbitrary frame is obtained by Lorentz boost


For a spin-1/2 particle at rest with spin along the z-axis, the spin vector is . This differs from the conventional normalization of s by a factor of the fermion mass m. Here we use the relativistic spinors normalized to 2E. In the extreme relativistic limit have s=Hk, where k is the lepton momentum and H is the lepton helicity.

zmˆs

Using trace theorems we obtain for leptonic tensor

Unpolarized lepton beam probes only the symmetric part of hadronic tensor


The Hadronic Tensor for Spin-1/2 TargetsUsing parity, time-reversal invariance, hermiticity and current conservation one can show that

Where the structure functions . and on depend and , , 2

2121 QggFF

Often another structure functions are used in the literature:


In elastic scattering there is a strong dependence on , and the elastic form factors fall o like a power of .Bjorken: in DIS the structure functions only depend on x, and must be independent of

The hadronic tensor is dimensionless

The structure functions are dimensionless functions of the Lorentz invariant variables

Scaling

222 and , QpqMp

It is conventional to write them as functions of

They can be written as dimensionless functions of the dimensionless variables

22 Q and 2 pqQx

22 MQ and x22 MQ

22 MQ

.2Q


For a longitudinally polarized lepton beam, the polarization is where is the lepton helicity.

The Cross-Section for Spin-1/2 Targets

Useful relation:

Contracting hadronic and leptonic tensors we get:

ksl 1


A target polarized along the incident beam direction:where for a target polarized parallel or antiparallel to the beam. in the evaluation of the cross-section

Longitudinally Polarized Target

zMh

s

1psh

Where the azimuthal angle has been integrated over since the cross-section is independent on


The polarization vector of a transversely polarized target can be chosen to point along the in the Lab frame. So, the azimuthal angle of scattered lepton is counted from that direction.Then, in this case we get:

Transversely Polarized Target

axisx

The structure functions g1 and g2 are equally important for a transversely polarized target, and so an experiment with a transversely polarized target can be used to determine g2, once g1 has been measured using a longitudinally polarized target.


We must construct only from the available 4-vectors, and , and the invariant tensors and . Thus we can write the most general structure in terms of the possible 6 tensors and correspondingform factors

Details of derivation for unpolarized DIS

Consider the general form of inelastic electron-proton scattering

W p q

g


Leptonic tensor is symmetric and we can ignore the antisymmetric terms in hadronic tensor for unpolarized DIS. Then conservation of theneutral current requires that , or, for arbitrary ,p q

Hence the coefficients of and in this equation mustseparately vanish

p q

Substituting back into the initial expression we have


In the laboratory frame we have the following relations between the kinematic quantities


Finally the definition of the cross section gives

inelastic scattering

Documents