Deep Inelastic Scattering

Deep inelastic scattering (DIS) is a high-energy scattering process in which a lepton transfers a large four-momentum to a hadron, producing a hadronic final state with a large invariant mass. Its principal targets are the proton, neutron, and atomic nuclei. DIS provides a central experimental method for investigating their quark and gluon content and testing quantum chromodynamics (QCD), the theory of the strong interaction. (pdg.lbl.gov)

Physical process and kinematics

The inclusive reaction is written

where $N$ is the target, $X$ denotes the hadronic final state, and the quantities in parentheses are four-momenta. The exchanged boson carries $q=k-k'$. In electromagnetic DIS it is a virtual photon; weak processes involve the W and Z bosons. (pdg.lbl.gov)

Using natural units, $c=\hbar=1$, the main invariant variables are

Here $Q^2>0$ measures the spacelike momentum transfer, $x$ is Bjorken $x$, $y$ is the inelasticity, and $W$ is the invariant mass of $X$. For a target of mass $M$,

“Deep” indicates large $Q^2$; “inelastic” indicates a hadronic mass well above the elastic and low-lying resonance regions. Large beam energy alone does not ensure DIS: the actual momentum transfer and final-state mass must also be large. (pdg.lbl.gov)

For negligible lepton masses, measurements in the target rest frame give

where $E$, $E'$, and $\theta$ are the incident energy, outgoing energy, and scattering angle. These relations allow the internal response of the target to be reconstructed from the scattered lepton. (slac.stanford.edu)

Structure functions and the parton model

Measured differential cross sections are expressed through structure functions, which parameterize the target’s response without initially assuming a microscopic model. Unpolarized electromagnetic DIS uses $F_1(x,Q^2)$ and $F_2(x,Q^2)$. Weak scattering additionally accesses $F_3$, associated with parity-violating contributions. (slac.stanford.edu)

In the parton model, the short-distance interaction is approximated as scattering from an individual constituent, or parton. Quarks, antiquarks, and gluons are described by parton distribution functions (PDFs). At leading order, $x$ is interpreted as the fraction of the target’s longitudinal momentum carried by the struck quark. Beyond leading order, Bjorken $x$ remains an observable kinematic variable but is not generally identical to the parton’s momentum fraction. (slac.stanford.edu)

For electromagnetic scattering in the leading-order, massless quark model,

where $e_f$ is the quark charge in units of the elementary charge. The Callan–Gross relation,

reflects scattering from spin-$\tfrac12$ constituents in this approximation. QCD radiation produces a nonzero longitudinal structure function $F_L=F_2-2xF_1$ when target-mass effects are neglected. (slac.stanford.edu)

Scaling and quantum chromodynamics

Bjorken scaling is the approximate independence of dimensionless structure functions from $Q^2$ at fixed $x$, in the limit of large momentum transfer and energy transfer. Its observation supported the interpretation of nucleons as containing effectively pointlike constituents rather than a smooth charge distribution alone. (slac.stanford.edu)

Scaling is not exact. In QCD, quarks emit gluons, and gluons split into quark–antiquark pairs or other gluons. These processes change the resolved parton distributions as the probing scale increases. The resulting logarithmic scaling violations are described by the DGLAP evolution equations. Asymptotic freedom makes perturbative calculations increasingly applicable at sufficiently large $Q^2$. (arxiv.org)

The theoretical description separates short-distance coefficient functions, calculable using perturbation theory, from long-distance PDFs:

The convolution integrates over parton momentum fractions. PDFs must be determined from measurements or other nonperturbative information; perturbative QCD predicts their scale evolution rather than their complete initial shapes. DIS therefore constrains both PDFs and the strong coupling, supplying essential inputs for predictions at the Large Hadron Collider. (arxiv.org)

Experimental forms and applications

Inclusive DIS sums over hadronic final states and primarily measures structure functions. Semi-inclusive DIS also identifies a produced hadron, providing additional sensitivity to quark flavor and transverse momentum. Its interpretation involves fragmentation functions describing how partons form observable hadrons through hadronization. (pdg.lbl.gov)

Experiments with polarized beams and targets measure spin-dependent distributions and structure functions such as $g_1$ and $g_2$. These investigate how parton polarization contributes to nucleon spin and provide access to quark–gluon correlations. Neutrino and antineutrino scattering supplies complementary flavor information through weak interactions. (jlab.org)

Scattering from nuclei tests whether bound nucleons have the same partonic response as free nucleons. The EMC effect denotes modifications of nuclear DIS structure functions relative to a simple collection of free protons and neutrons. Nuclear measurements must consequently distinguish nucleon structure from binding, motion, and other nuclear effects. (jlab.org)

Historical development

The SLAC–MIT experiments reported their first preliminary deep-inelastic electron–proton results in 1968. The unexpectedly substantial scattering rates and emerging scaling behavior provided decisive evidence for small, hard constituents within nucleons. Subsequent measurements helped establish the quark–parton description as a physical account of nucleon structure. (slac.stanford.edu)

Jerome I. Friedman, Henry W. Kendall, and Richard E. Taylor received the 1990 Nobel Prize in Physics for their pioneering investigations of deep-inelastic electron scattering on protons and bound neutrons. (nobelprize.org)

The HERA electron–proton collider extended DIS into a much broader kinematic range, including small $x$, and supplied precision measurements of proton structure and QCD dynamics. Its collider program began producing DIS results in 1992; the accelerator shut down at the end of June 2007. Its measurements remain important inputs to hadron-collider phenomenology. (www-h1.desy.de)

Limitations and interpretation

At moderate $Q^2$, target-mass corrections and higher-twist contributions complicate the leading-power description. At large $x$, the relation between $W^2$, $Q^2$, and $x$ can bring measurements close to the resonance region even when $Q^2$ is appreciable. Extracting PDFs there requires careful treatment of finite-mass and final-state effects. (arxiv.org)

At very small $x$, large logarithms of $1/x$ can make fixed-order calculations insufficient, motivating small-$x$ resummation. Their treatment affects extracted distributions and extrapolations outside the measured region. (arxiv.org)

DIS does not directly reveal a static three-dimensional image of a proton. Inclusive measurements determine combinations of momentum-dependent distributions, and different flavors or gluon contributions require complementary observables. Their extraction depends on theoretical approximations, experimental coverage, and, for nuclear targets, nuclear corrections. (jlab.org)

References

1. Structure Functionspdg.lbl.gov
2. The Theory of Deeply Inelastic Scatteringarxiv.org
3. The Nobel Prize in Physics 1990 – Illustrated presentationnobelprize.org
4. The Nobel Prize in Physics 1990nobelprize.org
5. Jefferson Lab Angular Momentum Collaborationjlab.org
6. Extraction of Deep Inelastic Cross Sections Using a 10.4 GeV Electron Beam and a Polarized Helium-3 Targetmisportal.jlab.org
7. Run Group L – 2025jlab.org
8. H1 publications (short listing)www-h1.desy.de
9. The end of HERAdesy.de
10. Collinear factorization for deep inelastic scattering structure functions at large Bjorken xBarxiv.org