A perturbative QCD study of dijets in p+Pb collisions at the LHC

TL;DR

This work evaluates perturbative QCD predictions for dijet production in proton+lead collisions at the LHC, focusing on how NLO corrections, scale choices, and free-proton PDF uncertainties compare to nuclear effects encoded in gluon modifications $R_G^A(x,Q^2)$. By analyzing the dijet rapidity distribution and the forward/backward yield ratio, the authors show that baseline uncertainties are small in the bulk region, and that these observables are highly sensitive to nuclear gluon modifications, enabling stringent tests of collinear factorization and constraints on nuclear PDFs. Differences among EPS09, DSSZ, and HKN07 are linked to distinct treatments of gluon modifications, especially at moderate-to-large $x$, suggesting that upcoming CMS data can discriminate between parametrizations. The study emphasizes symmetric rapidity acceptance as a route to minimizing experimental and theoretical baselines while enhancing sensitivity to $R_G^A(x,Q^2)$.

Abstract

Inspired by the recent measurements of the CMS collaboration, we report a QCD study of dijet production in proton+lead collisions at the LHC involving large-transverse-momentum jets, $p_T \gtrsim 100$ GeV. Examining the inherent uncertainties of the next-to-leading order perturbative QCD calculations and their sensitivity to the free proton parton distributions (PDFs), we observe a rather small, typically much less than 5% clearance for the shape of the dijet rapidity distribution within approximately 1.5 units around the midrapidity. Even a more stable observable is the ratio between the yields in the positive and negative dijet rapidity, for which the baseline uncertainty can be made negligible by imposing a symmetric jet rapidity acceptance. Both observables prove sensitive to the nuclear modifications of the gluon distributions, the corresponding uncertainties clearly exceeding the estimated baseline uncertainties from the free-proton PDFs and scale dependence. From a theoretical point of view, these observables are therefore very suitable for testing the validity of the collinear factorization and have a high potential to provide precision constraints for the nuclear PDFs.

Figures (7)

• Figure 1: The gluon nuclear modification factors from EPS09 (blue line with error band), DSSZ (green line with error bars) and HKN07 (purple dashed line) at $Q^2 = 10000 \, {\rm GeV}^2$. The approximate range probed by the CMS dijet measurements CMSprel is indicated by the thick black line.
• Figure 2: Upper panel: The absolute dijet spectrum at LO (dashed red line) and NLO (continuous blue line). The scale uncertainties are marked by the area enclosed by the dotted lines (LO), and the shaded band (NLO). The variable $\eta_{\rm dijet}$ is in the laboratory frame, and the vertical dotted line marks the location of the center-of-mass midrapidity. Lower panel: The ratio between the NLO and LO calculations (black line). Also shown are the relative CT10 PDF uncertainties (error bars) and the relative scale uncertainties in NLO (shaded band) and LO (band between dotted lines).
• Figure 3: Upper panel: The normalized dijet spectrum at LO (dashed red line) and NLO (continuous blue line). Lower panel: The ratio between the NLO and LO calculations (black line). Also shown are the relative CT10 PDF uncertainties (error bars) and the relative scale uncertainties in NLO (shaded band). The dotted lines are to guide the eye for a 2.5% relative uncertainty band.
• Figure 4: Upper panel: The forward-to-backward ratio at LO (dashed red line) and NLO (continuous blue line) as a function of the variable $\eta^*=|\eta_{\rm dijet}-\eta_{\rm shift} |$. Lower panel: The ratio between the NLO and LO calculations (black line). Also shown are the relative CT10 PDF uncertainties (error bars) and the relative scale uncertainties in NLO (shaded band).
• Figure 5: Upper panel: The normalized NLO dijet spectrum without nuclear effects in PDFs (orange dotted line) and with nuclear modifications from EPS09 (blue solid line), DSSZ (green dotted line), and HKN07 (purple dashed line). Lower panel: The normalized NLO dijet spectrum with nuclear effects divided by the corresponding calculation without the nuclear effects. The blue band corresponds to the EPS09 uncertainty, and the green hatched band is the DSSZ uncertainty range. The dotted lines mark again the 2.5% baseline uncertainty as in Fig. \ref{['Fig:Spectra22']}.