This classic 1995 paper from the New Muon Collaboration (NMC) at CERN represents a milestone in our understanding of the internal landscape of the proton and neutron. By firing high-energy muons at hydrogen and deuterium targets, the researchers mapped the "structure functions" that describe how quarks and gluons share the nucleon's momentum, providing a high-precision bridge between low-energy and high-energy physics.
The Challenge of Nucleon Structure
Protons and neutrons are not point-like particles; they are complex systems of quarks bound together by gluons. To understand the Strong Force (the "glue" of the universe), physicists use Deep Inelastic Scattering (DIS). In DIS, a high-energy lepton (like a muon) strikes a nucleon, and the resulting scattering pattern reveals the nucleon's internal structure.
The challenge in the mid-90s was consistency and coverage. Different experiments, such as SLAC at lower energies and BCDMS at higher energies, covered different "kinematic corners." There was a need for a high-precision experiment that could overlap these regions, resolve normalization discrepancies, and provide a unified picture of how the momentum distribution changes with the "hardness" of the collision (the $Q^2$ scale).
Experimental Methodology
The experiment (NMC-NA37) was conducted at the CERN SPS using a muon beam at four incident energies: 90, 120, 200, and 280 GeV. To minimize systematic errors, the team used two pairs of 3-meter-long targets (liquid hydrogen and liquid deuterium), periodically swapping their order in the beam to cancel out detector acceptance effects.
Extracting the structure function $F_2$ required correcting for detector resolution ("smearing") and higher-order electroweak effects. The authors used an iterative Monte Carlo approach, refining their model of the nucleon until it matched the raw data. Ultimately, they fitted a 15-parameter phenomenological function to their data combined with SLAC and BCDMS results to create a "global" description of $F_2$.
Reconciling Global Datasets
One of the paper's biggest contributions was determining the relative normalization shifts needed to make different experiments agree. By analyzing the overlap between datasets, the NMC team identified specific adjustments required to align the global data into a single, coherent curve.
As shown in the normalization table, the BCDMS data required a -2.3% shift and the NMC 90 GeV data a -2.2% shift for the proton measurements. These adjustments resolved long-standing discrepancies and allowed for the observation of "scaling violations"—the $Q^2$ dependence of $F_2$ that serves as a smoking gun for the existence of gluons.
Results: A Unified 15-Parameter Model
The measurements cover a wide kinematic range ($0.006 < x < 0.6$ and $0.5 < Q^2 < 75 \text{ GeV}^2$). The resulting 15-parameter fit provided a description of $F_2$ accurate to within a few percent over the range $0.006 < x < 0.9$.
This unified model showed excellent agreement with early collider results from HERA for $x \ge 0.01$, validating the transition from fixed-target to collider physics. The high statistical accuracy achieved meant that systematic uncertainties (ranging from 2% to 5%) became the dominant limiting factor, rather than raw data volume.
Legacy and Future Directions
While $x=0.006$ was deep for the time, the authors noted that even lower $x$ values—the "sea quark" and gluon-dominated region—would require the higher energies of the HERA collider. Additionally, the extraction of $F_2$ depended on $R$ (the ratio of longitudinal to transverse virtual photon cross-sections), which was taken from external parametrizations rather than being measured directly.
Despite these limitations, this work laid the essential groundwork for modern Parton Distribution Functions (PDFs). These functions remain indispensable today for making precise predictions at the Large Hadron Collider (LHC).
Summary updated