J/psi production in NRQCD: A global analysis of yield and polarization

TL;DR

The paper tests NRQCD factorization for J/psi production by performing a rigorous NLO global fit of the three color-octet LDMEs to 194 data points from hadro-, photo-, gamma-gamma, and e+e- experiments. It demonstrates that the CO LDMEs can describe a wide range of yield data and that the fitted values follow NRQCD’s expected scaling, while color-singlet contributions alone fall short. Using these LDMEs, it predicts J/psi polarization across production channels, finding partial consistency with some data but notable conflicts with Tevatron Run II measurements, and highlights ALICE data as more compatible. The results emphasize the potential of future LHC polarization measurements to test LDME universality and NRQCD factorization, while acknowledging unresolved feed-down and channel-specific tensions.

Abstract

We present a rigorous next-to-leading order analysis of J/psi yield and polarization within the factorization theorem of nonrelativistic QCD (NRQCD). To the orders considered, this framework depends on three free parameters, the color-octet long-distance matrix elements. We extract their values in a global fit to inclusive J/psi production data from various hadroproduction, photoproduction, two-photon scattering and electron-positron annihilation experiments. We show that this fit is constrained and stable and describes all data sufficiently well. We then make predictions for J/psi polarization in photo- and hadroproduction and compare them to the currently available data. As for photoproduction, HERA data is not precise enough to draw definite conclusions. But as for hadroproduction, CDF data measured at Tevatron run II is in strong conflict with NRQCD predictions. With early ALICE data being however compatible with NRQCD, the future, more precise polarization measurements at the LHC will thus have the potential to clearly confirm or dismiss LDME universality.

Figures (8)

• Figure 1: Plots a-t: Results of the global fit Butenschoen:2011yh compared to ALICE ALICEdata, ATLAS ATLASdata, Belle :2009nj, CDF Acosta:2004ywAbe:1997jz, CMS Khachatryan:2010yr, DELPHI Abdallah:2003du, LHCb Aaij:2011jh, PHENIX Adare:2009js, and ZEUS Chekanov:2002at data. The blue bands are the CSM predictions, the yellow bands include the CO contributions. The bands are constructed by variation of the renormalization, factorization and NRQCD scales.
• Figure 1: Plots u-z (continuation): Results of the global fit Butenschoen:2011yh compared to H1 Adloff:2002exAaron:2010gz data. The blue bands are the CSM predictions, the yellow bands include the CO contributions. The bands are constructed by variation of the renormalization, factorization and NRQCD scales.
• Figure 2: Decomposition of the NLO CS+CO curves from figures \ref{['fig:fitgraphs1']}c, q, r, and t into the contributions of individual intermediate $c\overline{c}$ states. The line coding is the same for all four graphs. Red curves mean negative values. Please note that these curves are the short distance cross sections already multiplied by the corresponding LDMEs.
• Figure 3: Figure a shows a decomposition of the NLO CS+CO curve of the two-photon-scattering results from plot \ref{['fig:fitgraphs1']}q into direct, single- and double-resolved photon contributions. Figures b and c show decompositions of the NLO CS+CO curves of the $p_T$ and $z$ distributions in photoproduction shown in figures \ref{['fig:fitgraphs1']}r and t into direct and resolved contributions.
• Figure 4: Predictions for the ATLAS measurement Aad:2011sp. Note that these data are not part of our global fit, since they became public after our global fit was finished. The bands are again constructed by variation of the renormalization, factorization and NRQCD scales. At very high $p_T$, it will be necessary to resum large logarithms $\log(p_T^2/m_{J/\psi}^2)$. For instance, at $p_T=40$ GeV, $\alpha_s\log(p_T^2/M_{J/\psi}^2)\approx0.7$.