EMC Effect for Partons

• The EMC effect for partons is the modification of nucleon PDFs within nuclei, observed as suppression in valence, sea, and gluon distributions at intermediate Bjorken-x.
• Modern analyses using LFHQCD, x-rescaling, and global nPDF fits reveal that the effect scales with short-range nucleon correlations and SU(6)-breaking effects.
• Experimental probes from DIS, Drell-Yan, and heavy-ion data establish flavor-dependent modifications, guiding predictions for nuclear QCD dynamics.

The EMC effect for partons refers to the modifications of nucleon parton distribution functions (PDFs) when nucleons are embedded in nuclear matter, as observed in high-precision deep-inelastic scattering (DIS) experiments. Initially discovered as a suppression of the nuclear structure function $F_2^A$ at intermediate Bjorken $x$ (relative to a naive sum of free-proton and free-neutron contributions), the EMC effect has evolved into a central probe of nonperturbative QCD dynamics and nucleon modification mechanisms in the nuclear environment, including valence, sea, and gluon partons. The modern understanding of the EMC effect incorporates a range of theoretical frameworks, connects directly to the physics of nucleon short-range correlations (SRCs), accommodates flavor dependence, and encompasses both quark and gluon channels.

1. Fundamental Decomposition and Universality

A rigorous decomposition of the nuclear structure function is established within the Light-Front Holographic QCD (LFHQCD) framework as follows (Kim et al., 2022):

$$F_2^A(x) = Z F_2^p(x) + (A-Z) F_2^n(x) + F_2^{\text{nuclear}}(x),$$

where $F_2^{\text{nuclear}}(x)$ encapsulates all nuclear medium effects. The LFHQCD model specifies that this nuclear piece factorizes into a universal function $U(x)$ (encoding the $x$-shape of the modification) and nuclear-dependent SU(6)-breaking coefficients $\delta r_p(A,Z)$ and $\delta r_n(A,Z)$:

$$F_2^{\text{nuclear}}(x) = \frac{5x}{3}\big[q_4(x) - q_3(x)\big] \left[Z\delta r_p(A,Z) + N\delta r_n(A,Z)\right],$$

$U(x) = \frac{5x}{3} [q_4(x) - q_3(x)].$

Here $q_3(x)$ and $q_4(x)$ are leading-twist valence LFHQCD distributions (three-quark and five-parton configurations), and $U(x)$ is genuinely nucleus-independent. The parameters $\delta r_{p,n}$ quantitatively express the breaking of SU(6) symmetry in the nuclear potential and track the prevalence of short-range nucleon-nucleon correlations (Kim et al., 2022).

2. Flavor and Partonic Species Dependence

2.1 Quarks and Antiquarks

The EMC suppression was historically observed for valence quarks at $0.3 \lesssim x \lesssim 0.7$, but recent Drell-Yan data reveal an analogous suppression for sea (antiquark) distributions. The nuclear Drell-Yan ratio

$$R_{\bar{q}}^A(x) \approx \frac{\sigma_{A}^{\text{DY}}}{A \sigma_{N}^{\text{DY}}},$$

at $x_A \approx 0.25$–0.45 exhibits a 3–5% monotonic suppression for heavy nuclei, incompatible with expectations from Fermi motion or simple energy-loss effects, and indicative of genuine medium modification of sea quark PDFs. This "antiquark EMC effect" appears to be slightly shifted to lower $x$ compared to the valence quark EMC effect, necessitating its inclusion in global nPDF fits (Alvioli et al., 2022).

2.2 Gluons

The gluon EMC effect is diagnosed through heavy-quark (e.g., charm) production in DIS at future electron-ion colliders:

$$R^A_g(x,Q^2) \equiv \frac{g^A(x,Q^2)}{A g^p(x,Q^2)},$$

where the suppression in $R^A_g$ at intermediate $x$ is directly analogous to the quark EMC effect. The reduced charm cross section $\sigma_{A, \text{red}}^{c\bar{c}}$ shows a matching suppression and is linearly correlated with sub-threshold $J/\psi$ photoproduction, which connects gluon EMC suppression with SRC observables (Wang et al., 2024, Hu et al., 17 Jan 2026).

2.3 Flavor Nonuniversality

Isospin-dependent mean fields in the nuclear medium produce flavor-dependent EMC effects. In the NJL model with self-consistent scalar and isovector mean-fields, up-quark distributions experience more significant suppression than down-quark distributions in neutron-rich nuclei (e.g., in Pb at $x\approx0.6$, $R_u\approx0.78$, $R_d\approx0.90$), observable via parity-violating DIS (Cloët et al., 2012).

3. QCD Mechanisms and Dynamical Origin

The EMC effect cannot be explained by single-nucleon Fermi motion or binding effects alone, as demonstrated using QCD sum rules and impulse approximation analysis (Frankfurt et al., 2012, Frankfurt et al., 2010). Instead, it necessarily requires non-nucleonic degrees of freedom:

• Suppression of Point-Like Configurations (PLC): Nuclear interactions deplete the probability $P_{\rm PLC}$ to find small-sized, high-$x$ three-quark configurations in bound nucleons, shifting wavefunction strength into larger, more partonic configurations, as encoded by SU(6)–breaking in the LFHQCD framework (Kim et al., 2022, Frankfurt et al., 2012).
• Short-Range Correlations (SRC): The empirical linear relation between the EMC slope $S_A \equiv -dR_A/dx$ and the SRC scale factor $a_2(A)$ (plateau value of inclusive $e$–$A$ cross section at $x>1$) strongly suggests that nucleon virtuality and two-nucleon SRCs are the primary drivers. This is reflected in analytic and convolution models where off-shell corrections tied to virtuality induce the entire observed EMC suppression for all partonic species (Sargsian, 2012, Hu et al., 17 Jan 2026, Hen et al., 2014).
• Photon and Kinematic Effects: For $x \lesssim 0.5$, Lorentz-boosted Coulomb fields generate equivalent photon partons that, together with exact $x$-kinematics, account for $1$–$3\%$ of the suppression (especially for heavy nuclei), leaving the genuinely hadronic modification as a residual (Frankfurt et al., 2010, Frankfurt et al., 2012).

4. Quantitative Modelling: Phenomenology and Global Fits

4.1 Fits and Evolution Equations

Global analyses at NNLO parameterize nuclear PDFs as

$f_i^A(x,Q_0^2) \equiv w_i(x,A,Z) f_i(x,Q_0^2),$

$$w_i(x,A,Z) = 1 + \left(1 - A^{-1/3}\right) \frac{a_i + b_i x + c_i x^2 + d_i x^3}{(1-x)^{\beta_i}},$$

with $a_i, b_i, c_i, d_i, \beta_i$ determined by fits to nuclear DIS and Drell-Yan data (Tehrani et al., 2014). NNLO corrections are essential for capturing the observed $Q^2$-dependence of $R_{\rm EMC}$ and for modeling gluon and sea quark channels.

4.2 x-Rescaling and Statistical Approaches

The $x$-rescaling model replaces $x$ with $x^{*} = \eta x$ in the free-nucleon structure function, with $\eta(A)$ growing linearly with the A-dependent binding energy, offering a compact one-parameter fit to EMC data for all stable nuclei (Canal et al., 2012).

Statistical models view nucleons as parton gases characterized by temperature parameters $T(A)$; the nuclear medium induces a downward shift in $T$, softening the PDFs and generating an EMC-type suppression for valence, sea, and gluon distributions (Yu, 2016, Shao et al., 2010).

5. SRC–EMC Correlation and Predictive Relations

A robust linear empirical correlation is identified between the EMC slope in DIS and the SRC scaling coefficient $a_2(A)$ obtained from $x > 1$ inclusive scattering (Sargsian, 2012, Hu et al., 17 Jan 2026):

$$S_A^q = -\frac{dR_A}{dx} = a_q \left[a_2(A) - 1\right],$$

$$S_A^g = -\frac{dR_A^{c\bar{c}}}{dx} = a_g\, (\sigma_A^{\text{sub}}/A) + b_g,$$

with $a_q \sim 0.05$–$0.09$ (quarks), $a_g \sim 0.04$–$0.05$ (gluons). These relations, validated by multiple nPDF fits, confirm that both quark and gluon EMC effects share the same A-scaling as the probability of SRC pairs, establishing a universal unified picture of partonic nuclear modification (Wang et al., 2024, Hu et al., 17 Jan 2026).

6. Experimental Manifestations and Practical Consequences

• Measurement channels: Inclusive and semi-inclusive DIS (providing $F_2^A$ and flavor-tagged ratios), Drell-Yan processes (sensitive to antiquarks), and heavy-quark production (probing gluons) are principal observables for the EMC effect at the parton level (Alvioli et al., 2022, Hen et al., 2014).
• Parity-violating DIS: Flavor-dependent nuclear effects can be extracted via parity-violating asymmetries, directly probing the flavor splitting in $u$/$d$ nuclear PDFs for $N \neq Z$ targets (Cloët et al., 2012).
• Global nPDF fits: Modern global fits (EPPS21, nNNPDF3.0, nCTEQ15HQ, TUJU21) include constraints from $R_{\text{EMC}}$, $a_2(A)$, Drell-Yan, and heavy-flavor data, accommodating quark, antiquark, and gluon EMC effects (Hu et al., 17 Jan 2026, Tehrani et al., 2014).
• Heavy-ion phenomenology: Nuanced implementation of the EMC effect in initial-state nPDFs is essential for the quantitative description of final-state observables, such as the nuclear modification factor $R_{AA}$ for heavy-flavor mesons at RHIC and LHC energies; omission of EMC-modified PDFs leads to significant deviations from experimental data (Mirjalili et al., 2021).

7. Open Questions and Future Directions

• Small $x$ region: The interplay between shadowing, antishadowing, and the onset of gluon saturation remains an active area, especially given negative gluon densities at NNLO, which may signal saturation or the QGP limit (Tehrani et al., 2014, Kotikov et al., 2018).
• Isospin and flavor structure: Further measurements (e.g., precise $^{48} \text{Ca}/^{40} \text{Ca}$ ratios, parity-violating DIS) are required to map out the isovector sector and distinguish $u$ and $d$ quark modifications (Cloët et al., 2012, Frankfurt et al., 2012).
• Equivalence of approaches: The quantitative equivalence or complementarity of the x-rescaling, SRC-convolution, and color-fluctuation models is under continued scrutiny, notably via tagged EMC and final-state nucleon correlation studies (Hen et al., 2014, Sargsian, 2012).
• Role of photons: For $x \lesssim 0.5$, the calculable photon EMC effect must be properly subtracted before attributing residual suppression to dynamical QCD mechanisms, especially in heavy nuclei (Frankfurt et al., 2010, Frankfurt et al., 2012).
• Gluonic EMC effect: Forthcoming high-luminosity DIS and photoproduction data (EIC, JLab12) will substantially sharpen our knowledge of gluon modifications and their connection to SRCs (Wang et al., 2024).

In summary, the EMC effect for partons is a multifaceted phenomenon—deeply connected to nuclear QCD dynamics, the structure of nucleon wavefunctions, and short-range nucleon-nucleon correlations—with a universal partonic modification function, quantitative SRC–EMC linearity, and explicit impact across both valence, sea, and gluonic sectors. Ongoing and future experiments, together with refined nPDF global fits, are expected to fully map out its flavor, spin, and A dependence for all partonic species.