Prompt vs Non-Prompt D0 Mesons

• Prompt and non-prompt D0 mesons are neutral open-charm meson subpopulations produced in high-energy collisions, differentiated by decay vertex displacement and production origin.
• The analysis employs high-resolution vertex detectors and advanced machine learning classifiers to disentangle contributions from direct charm hadronization and beauty hadron decays.
• Precise yield ratios and suppression patterns reveal key insights into QCD dynamics, heavy-quark energy loss, and the interplay of multi-partonic interactions in various collision systems.

Prompt and non-prompt $D^0$ mesons are distinct sub-populations of neutral open-charm mesons produced in high-energy hadronic collisions. Prompt $D^0$ mesons originate either via hadronization of directly produced charm quarks or feed-down from higher excited charm resonances and decay extremely close to the collision vertex. In contrast, non-prompt $D^0$ mesons arise from the weak decay of beauty hadrons ($B\to D^0+X$) and exhibit a significantly displaced decay vertex due to the longer $B$-hadron lifetime. The separation and precise measurement of these components provide critical probes for quantum chromodynamics (QCD), heavy-quark production dynamics, multi-partonic interactions, hadronization, and in-medium transport properties in both proton-proton and heavy-ion environments.

1. Production Mechanisms and Theoretical Framework

Prompt $D^0$ mesons are produced at the primary interaction vertex predominantly through leading-order QCD hard scatterings such as gluon-gluon fusion ($g + g \rightarrow c + \bar{c}$) and quark-antiquark annihilation ($q + \bar{q} \rightarrow c + \bar{c}$), with subsequent fragmentation of the charm quark to a $D^0$ meson or via feed-down from higher-mass open-charm hadrons. In collinear factorization:

$$\frac{d\sigma_{\mathrm{prompt}}}{dp_T} = \sum_{i,j} \int dx_1 dx_2 f_i(x_1, \mu_F) f_j(x_2, \mu_F) \frac{d\hat{\sigma}_{ij\rightarrow c\bar{c}}}{dp_{T,c}} \otimes D_{c\rightarrow D}(z)\,\delta(p_T - z p_{T,c})$$

Non-prompt $D^0$0 mesons are produced from weak decays of beauty hadrons, which themselves originate from $D^0$1-quark production via analogous hard-scattering processes ($D^0$2, $D^0$3) and subsequent fragmentation:

$D^0$4

where $D^0$5 and $D^0$6 are fragmentation functions, and $D^0$7 encodes the $D^0$8 decay kinematics (Radhakrishnan et al., 11 Jan 2026).

2. Experimental Separation and Analysis Methodologies

The separation of prompt and non-prompt $D^0$9 mesons exploits the distinct decay topologies arising from the disparate lifetimes of charm and beauty hadrons. The ALICE, CMS, and other LHC detectors utilize high-resolution silicon vertexing and tracking to reconstruct the $D^0$0 decay. Key observables include:

• (a) Vertex Displacement and Impact Parameter: Prompt $D^0$1 mesons have mean decay lengths of $D^0$2 μm, while non-prompt candidates from $D^0$3 decays have typical $D^0$4-hadron $D^0$5 μm (Collaboration, 2023, Collaboration, 2021, Goswami et al., 2024). Multi-variate classifiers, such as Boosted Decision Trees (BDT), are trained on simulated data to discriminate prompt and non-prompt using input variables including track impact parameters, decay-vertex displacement, pointing angle, and PID information.
• (b) Pseudoproper Time and Decay Lengths: Variables such as $D^0$6 and transverse pseudoproper decay length $D^0$7 provide further discrimination (Goswami et al., 2024).
• (c) Yield Extraction: By applying multiple classifier cuts, the raw candidate yields $D^0$8 are decomposed into prompt and non-prompt components using efficiency matrices. The final non-prompt fraction is given by $D^0$9 (Collaboration, 2023, Collaboration, 2021).
• (d) Machine Learning Approaches: XGBoost, CatBoost, and Random Forest classifiers trained on topological and kinematic inputs achieve $B\to D^0+X$0 purity/efficiency for prompt/non-prompt tagging on simulated data, with high fidelity in physical observables across $B\to D^0+X$1, $B\to D^0+X$2, and multiplicity bins (Goswami et al., 2024).

3. Transverse Momentum, Multiplicity, and Event Shape Dependence

Prompt and non-prompt $B\to D^0+X$3 yields and fractions exhibit characteristic dependencies on transverse momentum ($B\to D^0+X$4), charged-particle multiplicity, and event topology.

• $B\to D^0+X$5 Dependence: The non-prompt fraction $B\to D^0+X$6 rises monotonically with $B\to D^0+X$7, from $B\to D^0+X$8–$B\to D^0+X$9 at $B$0–$B$1 GeV/$B$2 to $B$3 above $B$4 GeV/$B$5 in inclusive (INEL$B$60) $B$7 samples at 13 TeV (Collaboration, 2023). The $B$8 ratio is $B$9 at $D^0$0 GeV/$D^0$1 and grows to $D^0$2 at $D^0$3 GeV/$D^0$4 (Goswami et al., 2024), with similar behaviors predicted by PYTHIA 8 and observed in ALICE data at various energies (Collaboration, 2021).
• Multiplicity Dependence: $D^0$5 shows no significant change with multiplicity, remaining constant within $D^0$6 (with $D^0$7 the double ratio of non-prompt fractions between multiplicity classes) (Collaboration, 2023). However, self-normalized non-prompt $D^0$8 yields $D^0$9 rise faster than linearly with normalized charged-particle multiplicity $g + g \rightarrow c + \bar{c}$0, especially at high $g + g \rightarrow c + \bar{c}$1 and collision energy, due to the strong sensitivity of beauty production to multiple partonic interactions (MPI) (Goswami et al., 2024, Radhakrishnan et al., 11 Jan 2026).
• Event-Shape Engineering: Non-prompt $g + g \rightarrow c + \bar{c}$2 mesons demonstrate strong correlation with the hardest partonic scatter (high $g + g \rightarrow c + \bar{c}$3) and little dependence on late-stage color reconnection or event spherocity, reflecting their origin fixed by the primary hard process. In contrast, prompt $g + g \rightarrow c + \bar{c}$4 mesons receive feed-down from semi-hard processes and are more sensitive to color reconnection effects and event isotropy (Radhakrishnan et al., 11 Jan 2026).

4. Nuclear Modification and Collectivity in Heavy-Ion Collisions

In heavy-ion collisions, prompt and non-prompt $g + g \rightarrow c + \bar{c}$5 mesons serve as mass-differentiated probes of parton energy loss and quark-gluon plasma (QGP) transport properties.

• Suppression Patterns ($g + g \rightarrow c + \bar{c}$6): Non-prompt $g + g \rightarrow c + \bar{c}$7 $g + g \rightarrow c + \bar{c}$8 is consistently higher than that for prompt $g + g \rightarrow c + \bar{c}$9 and charged hadrons for $q + \bar{q} \rightarrow c + \bar{c}$0–$q + \bar{q} \rightarrow c + \bar{c}$1 GeV/$q + \bar{q} \rightarrow c + \bar{c}$2, expressing the dead-cone effect and reduced in-medium coupling for bottom quarks. For example, $q + \bar{q} \rightarrow c + \bar{c}$3(non-prompt $q + \bar{q} \rightarrow c + \bar{c}$4) $q + \bar{q} \rightarrow c + \bar{c}$5 0.25 (2–4 GeV/$q + \bar{q} \rightarrow c + \bar{c}$6), rising to $q + \bar{q} \rightarrow c + \bar{c}$7 0.80 (14–30 GeV/$q + \bar{q} \rightarrow c + \bar{c}$8) in central PbPb collisions, while prompt $q + \bar{q} \rightarrow c + \bar{c}$9 and charged hadron $D^0$0 remain lower (Collaboration, 2018, Xing et al., 2024). The $D^0$1 yield ratio increases with $D^0$2, with differences between $D^0$3 and PbPb most pronounced at low $D^0$4 due to stronger beauty suppression (Collaboration, 2018).
• Elliptic and Triangular Flow ($D^0$5, $D^0$6): Non-prompt $D^0$7 mesons display significant but smaller $D^0$8 and $D^0$9 than their prompt counterparts, confirming reduced thermalization and weaker collective coupling for beauty quarks. Typical values for $$\frac{d\sigma_{\mathrm{prompt}}}{dp_T} = \sum_{i,j} \int dx_1 dx_2 f_i(x_1, \mu_F) f_j(x_2, \mu_F) \frac{d\hat{\sigma}_{ij\rightarrow c\bar{c}}}{dp_{T,c}} \otimes D_{c\rightarrow D}(z)\,\delta(p_T - z p_{T,c})$$0(non-prompt $$\frac{d\sigma_{\mathrm{prompt}}}{dp_T} = \sum_{i,j} \int dx_1 dx_2 f_i(x_1, \mu_F) f_j(x_2, \mu_F) \frac{d\hat{\sigma}_{ij\rightarrow c\bar{c}}}{dp_{T,c}} \otimes D_{c\rightarrow D}(z)\,\delta(p_T - z p_{T,c})$$1) are 0.02–0.07 across $$\frac{d\sigma_{\mathrm{prompt}}}{dp_T} = \sum_{i,j} \int dx_1 dx_2 f_i(x_1, \mu_F) f_j(x_2, \mu_F) \frac{d\hat{\sigma}_{ij\rightarrow c\bar{c}}}{dp_{T,c}} \otimes D_{c\rightarrow D}(z)\,\delta(p_T - z p_{T,c})$$2 and centrality, with prompt $$\frac{d\sigma_{\mathrm{prompt}}}{dp_T} = \sum_{i,j} \int dx_1 dx_2 f_i(x_1, \mu_F) f_j(x_2, \mu_F) \frac{d\hat{\sigma}_{ij\rightarrow c\bar{c}}}{dp_{T,c}} \otimes D_{c\rightarrow D}(z)\,\delta(p_T - z p_{T,c})$$3 $$\frac{d\sigma_{\mathrm{prompt}}}{dp_T} = \sum_{i,j} \int dx_1 dx_2 f_i(x_1, \mu_F) f_j(x_2, \mu_F) \frac{d\hat{\sigma}_{ij\rightarrow c\bar{c}}}{dp_{T,c}} \otimes D_{c\rightarrow D}(z)\,\delta(p_T - z p_{T,c})$$4 reaching up to $$\frac{d\sigma_{\mathrm{prompt}}}{dp_T} = \sum_{i,j} \int dx_1 dx_2 f_i(x_1, \mu_F) f_j(x_2, \mu_F) \frac{d\hat{\sigma}_{ij\rightarrow c\bar{c}}}{dp_{T,c}} \otimes D_{c\rightarrow D}(z)\,\delta(p_T - z p_{T,c})$$5. The mass ordering $$\frac{d\sigma_{\mathrm{prompt}}}{dp_T} = \sum_{i,j} \int dx_1 dx_2 f_i(x_1, \mu_F) f_j(x_2, \mu_F) \frac{d\hat{\sigma}_{ij\rightarrow c\bar{c}}}{dp_{T,c}} \otimes D_{c\rightarrow D}(z)\,\delta(p_T - z p_{T,c})$$6 is observed in both ALICE and CMS, aligning with mechanistic expectations from Langevin transport and Boltzmann models (Collaboration, 2023, Collaboration, 2022, Xing et al., 2024).
• Theoretical Models: Heavy-quark energy loss calculations including mass-dependent drag and diffusion coefficients (TAMU, LBT, PHSD, CUJET, EPOS) describe the overall features of $$\frac{d\sigma_{\mathrm{prompt}}}{dp_T} = \sum_{i,j} \int dx_1 dx_2 f_i(x_1, \mu_F) f_j(x_2, \mu_F) \frac{d\hat{\sigma}_{ij\rightarrow c\bar{c}}}{dp_{T,c}} \otimes D_{c\rightarrow D}(z)\,\delta(p_T - z p_{T,c})$$7 and $$\frac{d\sigma_{\mathrm{prompt}}}{dp_T} = \sum_{i,j} \int dx_1 dx_2 f_i(x_1, \mu_F) f_j(x_2, \mu_F) \frac{d\hat{\sigma}_{ij\rightarrow c\bar{c}}}{dp_{T,c}} \otimes D_{c\rightarrow D}(z)\,\delta(p_T - z p_{T,c})$$8 for prompt and non-prompt $$\frac{d\sigma_{\mathrm{prompt}}}{dp_T} = \sum_{i,j} \int dx_1 dx_2 f_i(x_1, \mu_F) f_j(x_2, \mu_F) \frac{d\hat{\sigma}_{ij\rightarrow c\bar{c}}}{dp_{T,c}} \otimes D_{c\rightarrow D}(z)\,\delta(p_T - z p_{T,c})$$9, although in the low-$D^0$00 range non-prompt suppression can be stronger than standard models predict, possibly implying enhanced collisional drag or altered $D^0$01-hadron chemistry owing to coalescence (Collaboration, 2018, Collaboration, 2022, Xing et al., 2024).

5. Cross Sections, Yield Ratios, and Model Comparisons

Precise differential and total cross sections for prompt and non-prompt $D^0$02 mesons have been extracted at multiple energies.

$D^0$03 (GeV/$D^0$04)	$D^0$05 Prompt ($D^0$06b/(GeV/$D^0$07))	$D^0$08 Non-prompt ($D^0$09b/(GeV/$D^0$10))
1–2	$D^0$11	$D^0$12
4–6	$D^0$13	$D^0$14
8–12	$D^0$15	$D^0$16
16–24	$D^0$17	$D^0$18

In $D^0$19 for $D^0$20 at 5.02 TeV, the $D^0$21-integrated visible non-prompt $D^0$22 cross section is $D^0$23 μb (Collaboration, 2021). The $D^0$24 production cross section per rapidity unit at midrapidity, extracted via non-prompt $D^0$25 yields, is $D^0$26b, consistent with FONLL pQCD predictions (Collaboration, 2021). The non-prompt/prompt $D^0$27 yield ratio increases from $D^0$28 at $D^0$29 GeV/$D^0$30 to $D^0$31 at $D^0$32 GeV/$D^0$33.

Model comparisons:

• PYTHIA 8, especially with Colour Reconnection beyond Leading Colour (CR-BLC) or Colour Ropes, reproduces qualitative $D^0$34 trends but overestimates absolute yields by $D^0$35–$D^0$36, and predicts a slight multiplicity dependence disfavored by data (Collaboration, 2023, Radhakrishnan et al., 11 Jan 2026).
• EPOS 3/4 underpredict $D^0$37 and predict stronger multiplicity dependence than observed (Collaboration, 2023).
• CGC calculations with three-pomeron fusion are compatible with the observed double ratios (Collaboration, 2023).

6. Implications for QCD and Heavy-Flavor Dynamics

Simultaneous measurements of prompt and non-prompt $D^0$38 mesons constrain heavy-quark fragmentation functions, hadronization mechanisms, and the mass-dependence of parton diffusion and energy loss. The weak multiplicity dependence of $D^0$39 at midrapidity indicates similar multi-parton and hadronization dynamics for charm and beauty in $D^0$40 collisions, disfavoring scenarios of strong enhancement in beauty-baryon over beauty-meson yields at high multiplicity (Collaboration, 2023, Radhakrishnan et al., 11 Jan 2026). The observed hierarchy $D^0$41(charged)$D^0$42 $D^0$43(prompt $D^0$44)$D^0$45 $D^0$46(non-prompt $D^0$47) and $D^0$48(prompt)$D^0$49 $D^0$50(non-prompt) in heavy-ion collisions quantitatively embody color coherence and the dead-cone effect, providing direct experimental access to the bottom-quark transport coefficient $D^0$51 (Collaboration, 2022, Xing et al., 2024).

These measurements, enabled by advances in experimental reconstruction and machine learning, underpin precision tests of QCD production and non-perturbative dynamics in both elementary and nuclear systems. They also provide benchmarks for future, more differential extractions of heavy-quark transport parameters and heavy-flavor hadronization in the high-luminosity era.