Intrinsic Charm in Hadrons

• Intrinsic charm components are nonperturbative c-c̄ pairs in higher Fock states of hadrons, distinct from perturbative charm from gluon splitting.
• Modeling approaches like BHPS, sea-like, and meson–baryon cloud models parameterize intrinsic charm with fits from global PDF analyses and experimental data.
• Phenomenological evidence from deep inelastic scattering, LHC forward charm production, and atmospheric neutrino flux studies highlights its role in high-energy observables.

Intrinsic charm components refer to nonperturbative charm–anticharm ($c\bar c$) pairs present as higher Fock–state admixtures in the wavefunction of the nucleon or other hadrons, distinct from the perturbatively generated (extrinsic) charm arising from gluon splitting. These components, predicted by QCD but absent in simple three–quark models, encode long-distance aspects of QCD dynamics and can have observable phenomenological consequences in hadronic scattering, parton distribution functions (PDFs), and high-energy collider processes.

1. Theoretical Foundations of Intrinsic Charm

Quantum Chromodynamics (QCD) implies that the proton wavefunction includes configurations beyond its minimal valence content. In light-cone quantization, the nucleon state expands as a superposition of Fock states: $$|\mathrm{p}\rangle = a\,|uud\rangle + b\,|uudg\rangle + c\,|uudc\bar{c}\rangle + \dots$$ The $|uud\,c\bar{c}\rangle$ component is identified as "intrinsic charm” (IC), distinct from "extrinsic” perturbative $g\to c\bar c$ pairs produced via DGLAP evolution or hard scattering. The conceptual basis for intrinsic heavy quarks was established by Brodsky–Hoyer–Peterson–Sakai (BHPS) who derived the probability of finding such Fock states using old-fashioned light-cone Hamiltonian perturbation theory (Lykasov et al., 2012, Hobbs, 2016). The normalization of the intrinsic-charm content is typically characterized by the probability $P_{c\bar c} = \int_0^1 dx\,f_c^{\rm IC}(x)$ (with $f_c^{\rm IC}(x)$ the intrinsic charm distribution at a boundary scale $\mu_0 \sim m_c$).

Non-perturbative mechanisms such as meson–baryon fluctuations, chiral models, and light-cone wavefunction analyses support the existence of a small but nonzero IC component. In the meson–baryon cloud approach, the nucleon can fluctuate into charmed baryons and $D$ or $D^*$ mesons, generating asymmetric $c$ and $$|\mathrm{p}\rangle = a\,|uud\rangle + b\,|uudg\rangle + c\,|uudc\bar{c}\rangle + \dots$$0 distributions (Hobbs et al., 2013, Goncalves et al., 2024).

2. Modeling and Parameterization

Intrinsic charm is typically modeled as an additive non-perturbative term in the charm PDF: $$|\mathrm{p}\rangle = a\,|uud\rangle + b\,|uudg\rangle + c\,|uudc\bar{c}\rangle + \dots$$1 The canonical BHPS distribution at scale $$|\mathrm{p}\rangle = a\,|uud\rangle + b\,|uudg\rangle + c\,|uudc\bar{c}\rangle + \dots$$2 is: $$|\mathrm{p}\rangle = a\,|uud\rangle + b\,|uudg\rangle + c\,|uudc\bar{c}\rangle + \dots$$3 where $$|\mathrm{p}\rangle = a\,|uud\rangle + b\,|uudg\rangle + c\,|uudc\bar{c}\rangle + \dots$$4 is fixed by $$|\mathrm{p}\rangle = a\,|uud\rangle + b\,|uudg\rangle + c\,|uudc\bar{c}\rangle + \dots$$5. Other parameterizations include "sea-like" models, which take $$|\mathrm{p}\rangle = a\,|uud\rangle + b\,|uudg\rangle + c\,|uudc\bar{c}\rangle + \dots$$6 (Bailas et al., 2015, Hou et al., 2017). In meson–baryon models, convolution formulas account for the splitting of the proton into charmed baryons and mesons and yield $$|\mathrm{p}\rangle = a\,|uud\rangle + b\,|uudg\rangle + c\,|uudc\bar{c}\rangle + \dots$$7 at moderate $$|\mathrm{p}\rangle = a\,|uud\rangle + b\,|uudg\rangle + c\,|uudc\bar{c}\rangle + \dots$$8 (Hobbs et al., 2013, Goncalves et al., 2024, Ball et al., 2023).

Neural-network-based PDF fits, as in the NNPDF4.0 framework, incorporate a fully flexible fitted charm PDF that captures both the perturbative and intrinsic components without assuming a specific functional form—this flexibility enables a robust extraction of IC from experimental data (Ball et al., 2022, Rottoli, 2016).

3. Phenomenological Evidence and Global Fits

Multiple global PDF fits and phenomenological analyses, incorporating inclusive and semi-inclusive deep-inelastic scattering (DIS), Drell-Yan production, W/Z+c-jet, and heavy flavor hadron production data, have established constraints or evidence for IC:

• Statistically significant evidence: The NNPDF4.0 analysis finds a valence-like $$|\mathrm{p}\rangle = a\,|uud\rangle + b\,|uudg\rangle + c\,|uudc\bar{c}\rangle + \dots$$9, peaked at $|uud\,c\bar{c}\rangle$0 and vanishing below $|uud\,c\bar{c}\rangle$1, with a momentum fraction

$|uud\,c\bar{c}\rangle$2

and statistical significance at or above $|uud\,c\bar{c}\rangle$3 (Ball et al., 2022).

• Earlier and parallel results: NNPDF3IC finds $|uud\,c\bar{c}\rangle$4 (with EMC), very similar to the current NNPDF4.0 results (Rottoli, 2016).
• CT14/CT14HERA2: NNLO global fits tolerate up to $|uud\,c\bar{c}\rangle$5 (BHPS) at 90% C.L., with a mild preference for $|uud\,c\bar{c}\rangle$6 (Hou et al., 2017).
• Direct QCD sum rule calculations: Analytic results from interpolating current techniques yield $|uud\,c\bar{c}\rangle$7, within the ballpark of phenomenological global-fit values (Olamaei et al., 2023).
• Meson–baryon models & asymmetry: Convolution calculations predict a negative $|uud\,c\bar{c}\rangle$8 asymmetry at moderate-to-large $|uud\,c\bar{c}\rangle$9 and a momentum fraction $g\to c\bar c$0 (Hobbs et al., 2013). Modern global fits now include fits for $g\to c\bar c$1 (Ball et al., 2023, Goncalves et al., 2024).

Empirically, forward open charm production in hadronic collisions, forward $g\to c\bar c$2 or $g\to c\bar c$3 production at the LHC, and associated $g\to c\bar c$4 or $g\to c\bar c$5 production provide sensitive probes of the large-$g\to c\bar c$6 charm PDF (Lykasov et al., 2012, Boettcher et al., 2015, Bailas et al., 2015).

4. Phenomenological Impact and Observables

The presence of IC alters key observables, especially at high $g\to c\bar c$7:

• Structure functions: The heavy structure function $g\to c\bar c$8 is sensitive to $g\to c\bar c$9 at large $P_{c\bar c} = \int_0^1 dx\,f_c^{\rm IC}(x)$0. Inclusion of an IC component typically produces a pronounced bump at $P_{c\bar c} = \int_0^1 dx\,f_c^{\rm IC}(x)$1 and improves the description of historical EMC $P_{c\bar c} = \int_0^1 dx\,f_c^{\rm IC}(x)$2 data for $P_{c\bar c} = \int_0^1 dx\,f_c^{\rm IC}(x)$3 (Rottoli, 2016, Abdolmaleki et al., 2019).
• LHC observables: The cross section for $P_{c\bar c} = \int_0^1 dx\,f_c^{\rm IC}(x)$4 production at large rapidity, $P_{c\bar c} = \int_0^1 dx\,f_c^{\rm IC}(x)$5, is highly sensitive to the presence and shape of IC. For $P_{c\bar c} = \int_0^1 dx\,f_c^{\rm IC}(x)$6 (BHPS2), this ratio can be enhanced by a factor of up to $P_{c\bar c} = \int_0^1 dx\,f_c^{\rm IC}(x)$7 at $P_{c\bar c} = \int_0^1 dx\,f_c^{\rm IC}(x)$8 (Bailas et al., 2015, Boettcher et al., 2015).
• Forward production: In forward $P_{c\bar c} = \int_0^1 dx\,f_c^{\rm IC}(x)$9 collisions, D-meson yields at high pseudorapidity are roughly doubled at $f_c^{\rm IC}(x)$0 when including a $f_c^{\rm IC}(x)$1 IC (CTEQ66c) compared to no-IC (CTEQ66) (Lykasov et al., 2012).
• Prompt atmospheric neutrinos: The prompt neutrino flux at high energy (as measured by IceCube) is enhanced by up to an order of magnitude at $f_c^{\rm IC}(x)$2 GeV for $f_c^{\rm IC}(x)$3. This constrains $f_c^{\rm IC}(x)$4 from the requirement not to overshoot the observed flux (Laha et al., 2016, Maciula et al., 2021).
• Charm–anticharm asymmetry: Asymmetric IC models, e.g., meson–baryon cloud and NNPDF4.0 fitted-charm, predict $f_c^{\rm IC}(x)$5. The magnitude and sign of the predicted $f_c^{\rm IC}(x)$6 asymmetry in forward $f_c^{\rm IC}(x)$7 fixed-target collisions is sensitive to the model details; current data indicate that existing IC models alone, or with recombination, do not fully explain the observed large-$f_c^{\rm IC}(x)$8 asymmetries (Goncalves et al., 2024, Ball et al., 2023).

The table below summarizes characteristic model features and current empirical constraints:

Model/Approach	IC Momentum Fraction	$f_c^{\rm IC}(x)$9-Shape
BHPS (light-cone)	$\mu_0 \sim m_c$0 (input/model)	Valence-like, peak at $\mu_0 \sim m_c$1
Sea-like (proportional to $\mu_0 \sim m_c$2)	$\mu_0 \sim m_c$3 (input/model)	Soft, peaks at $\mu_0 \sim m_c$4
NNPDF4.0 (fitted)	$\mu_0 \sim m_c$5 (data, $\mu_0 \sim m_c$6)	Data-driven, valence-like
Meson–baryon cloud (MBM)	$\mu_0 \sim m_c$7 (fit to $\mu_0 \sim m_c$8 data)	$\mu_0 \sim m_c$9, asymmetric peak at $D$0
QCD sum rule (analytic)	$D$1	--
Constraints from IceCube	$D$2	Consistent with BHPS allowed

5. Theoretical Status and QCD Factorization

Within general-mass, variable-flavor-number schemes (VFNS) for heavy quarks, intrinsic charm formally appears as a scale-independent boundary condition for the charm PDF at $D$3 (Ball et al., 2015). In the absence of IC, FONLL and S-ACOT schemes reduce to the same formula; if IC is present, the ACOT and FONLL prescriptions recover the full cross-section including its contributions at leading power in $D$4, with nonzero $D$5. Theoretically, IC corresponds to "twist-4" proton matrix elements, suppressed by powers of $D$6 but not vanishing at large $D$7 (Hou et al., 2017, Hobbs, 2016).

The sum rules for PDF normalization remain satisfied due to compensating small adjustments in the gluon and light-sea PDFs when IC is included.

6. Experimental and Observational Probes

Robust constraints and evidence for IC rely on multiple experimental strategies:

• DIS structure functions $D$8: High-$D$9 charm structure function data remain essential. The EMC data at $D^*$0 provide the strongest evidence for nonzero IC, though newer global fits caution that systematic uncertainties and tensions with HERA data limit their impact (Rottoli, 2016, Hobbs, 2016).
• Forward open-charm and charmed baryon production: Enhancement of D-meson and $D^*$1, $D^*$2 yields at high Feynman-$D^*$3 or rapidity is a clean IC signature, especially in kinematic regimes where the $D^*$4 channel dominates (Lykasov et al., 2012, Kopeliovich et al., 2010).
• $D^*$5 and $D^*$6 production at the LHC, especially LHCb: High-rapidity and high-$D^*$7 $D^*$8 cross sections, and the ratios $D^*$9, are direct probes of the IC contribution at $c$0 (Bailas et al., 2015, Boettcher et al., 2015, Ball et al., 2022).
• Prompt atmospheric neutrino flux in IceCube: The forward production of charm in cosmic-ray collisions leads to an increased prompt neutrino background, which is highly sensitive to the large-$c$1 IC contribution in the proton (Laha et al., 2016, Maciula et al., 2021).
• $c$2 and charm–anticharm production asymmetries: Measurement of $c$3-meson production asymmetries in fixed-target and collider environments informs $c$4 models and the valence IC PDF (Goncalves et al., 2024, Ball et al., 2023).

Continued and future measurements, especially at the Electron-Ion Collider (with flavor-tagged structure functions) and at forward LHC and fixed-target programs, are expected to further pin down the normalization and $c$5-shape of the intrinsic charm component (Boettcher et al., 2015, Ball et al., 2022).

7. Open Challenges and Outlook

While significant progress has been made in establishing the existence and characterizing the properties of intrinsic charm, several challenges remain:

• Normalization and Uncertainties: The precise normalization ($c$6) is still subject to uncertainties from experimental systematics, the treatment of higher-order QCD corrections, and the variety of hadronic and nuclear corrections needed in the interpretation of data (Hou et al., 2017, Abdolmaleki et al., 2019).
• Model Discrimination: Disentangling valence-like from sea-like IC, and distinguishing among BHPS, meson–baryon, and data-driven fitted shapes, requires finer binning and more differential measurements at high $c$7 (Boettcher et al., 2015, Hobbs et al., 2013, Goncalves et al., 2024).
• Charm–anticharm asymmetry: While asymmetric models predict $c$8, fully describing observed $c$9 asymmetries at large $$|\mathrm{p}\rangle = a\,|uud\rangle + b\,|uudg\rangle + c\,|uudc\bar{c}\rangle + \dots$$00 may require improved initial-state or final-state physics beyond standard IC models (Goncalves et al., 2024, Ball et al., 2023).
• Impact on SM precision and BSM physics: A sub-percent-level intrinsic charm alters predictions for $$|\mathrm{p}\rangle = a\,|uud\rangle + b\,|uudg\rangle + c\,|uudc\bar{c}\rangle + \dots$$01, $$|\mathrm{p}\rangle = a\,|uud\rangle + b\,|uudg\rangle + c\,|uudc\bar{c}\rangle + \dots$$02, Higgs+c, and heavy quarkonium at present and future colliders. Accurate knowledge of IC is thus essential for both QCD and possible new physics extraction (Bailas et al., 2015, Boettcher et al., 2015).
• Connections to lattice QCD and sum rule calculations: QCD sum rule calculations yield results compatible with global-fit extractions, but lattice determinations of the charm content in the nucleon (charmness–sigma term) currently have large uncertainties (Olamaei et al., 2023, Duan et al., 2016).
• Experimental confirmation: Unambiguous identification demands precision mapping of $$|\mathrm{p}\rangle = a\,|uud\rangle + b\,|uudg\rangle + c\,|uudc\bar{c}\rangle + \dots$$03 at large $$|\mathrm{p}\rangle = a\,|uud\rangle + b\,|uudg\rangle + c\,|uudc\bar{c}\rangle + \dots$$04, high-statistics measurement of forward charm production and $$|\mathrm{p}\rangle = a\,|uud\rangle + b\,|uudg\rangle + c\,|uudc\bar{c}\rangle + \dots$$05 at LHCb and future EIC data with flavor tagging (Boettcher et al., 2015, Ball et al., 2022, Ball et al., 2023).

Intrinsic charm thus constitutes a nonperturbative, experimentally accessible aspect of nucleon structure, now increasingly constrained by multi-process global analyses, and remains an active subject at the intersection of QCD theory, collider phenomenology, and astroparticle physics.