Jet quenching for hadron-tagged jets in $pA$ collisions

Abstract

We calculate the medium modification factor $I_{pA}$ for 5.02 TeV $p$+Pb collisions. We use the Monte-Carlo Glauber model to determine the parameters of the quark-gluon plasma fireball in $pA$ jet events. Our calculations show that the jet quenching effect for $I_{pA}$ turns out to be rather small. We have found that the theoretical $I_{pA}$ as a function of the underlying event charged multiplicity density, within errors, agrees with data from ALICE ALICE_Ipp_PLB for 5.02 TeV $p$+Pb collisions. However, the experimental errors are too large to draw a firm conclusion on the possible presence of jet quenching.

Introduction. The transverse flow effects and the strong suppression of high-$p_{T}$ hadron spectra (jet quenching) observed in $AA$ collisions at RHIC and the LHC are strong evidence for formation of the quark-gluon plasma (QGP) (for reviews see, e.g., hydro2 ; hydro3 ; JQ_rev ). The results of hydrodynamic modeling of the flow effects hydro2 ; hydro3 and of the jet quenching analyses of the heavy ion data on the nuclear modification factor $R_{AA}$ (see, e.g., CUJET3 ; Armesto_LHC ; Z_hl ) are consistent with the onset of the collective/hydrodynamic behavior of the QCD matter at the proper time $\tau_{0}\sim 0.5-1$ fm. Although the dynamics of the QCD matter in $AA$ collisions was under active investigation in the last years for the weak and strong coupling scenarios (for reviews see, e.g., Teaney1 ; Strickland1 ; chesler1 ), the mechanism of the early hydrodynamization of the QGP is not well understood yet. The observation of the QGP formation in $AA$ collisions has sparked new interest in the idea of the QGP formation for small systems Shuryak_QGP . The analysis of Chesler2 in the scenario of the strongly coupled QGP shows that even the smallest droplet of the QCD matter produced in $pp/pA$ collisions can be described within hydrodynamics. In Spalinski it was argued that for $pp/pA$ collisions the lower bound for applicability of hydrodynamical description is $dN_{ch}/d\eta\sim 3$. This is close to the estimate of RMK1609 .

Experimentally, the formation of a mini QGP in $pp/pA$ collisions is supported by the observation of the ridge effect CMS_ridge ; ATLAS_mbias ; ATLAS_ridgepA ; ALICE_ridgepA in $pp/pA$ collisions at the LHC energies and by the steep growth of the midrapidity strange particle production at charged multiplicity $dN_{ch}/d\eta\mathrel{\hbox to0.0pt{\lower 4.0pt\hbox{\hskip 1.0pt$\sim$}\hss}\raise 1.0pt\hbox{$>$}}5$ ALICE_strange . The earlier analysis Camp1 of $\langle p_{T}\rangle$ as a function of multiplicity, employing Van Hove’s arguments VH also supports the onset of QGP regime at such charged multiplicity density. There were suggested alternative non-hydrodynamical explanations of the ridge effect in $pp/pA$ collisions Kovner_ridge1 ; Dumitru_ridge1 ; ridge_CGC due to the initial state parton effects. However, these models do not explain the anomalous variation with the charged multiplicity density of the midrapidity strange particle production and of $\langle p_{T}\rangle$.

Besides the generation of the flow effects, the QGP formation in $pp/pA$ collisions should lead to jet modification due to parton energy loss in the QGP fireball. It is important that the typical charged multiplicity of soft (underlying event (UE)) hadrons in jet events is bigger than the ordinary minimum bias multiplicity by a factor of $\sim 2-2.5$ Field . For the LHC energies the typical midrapidity charged multiplicity in $pp$ jet events $dN_{ch}/d\eta\sim 10-15$. One of the possible experimental methods to probe jet quenching in the small size QGP produced in $pp/pA$ collisions is investigation of the UE multiplicity dependence of the jet fragmentation functions (FFs) for photon/hadron tagged jets Z_pp_PRL described by the modification factors $I_{pp/pA}$. Formally, $I_{pp,pA}$ can be defined as the ratio of the per-trigger particle ($h^{t}$) yield of the associated hadron ($h^{a}$) production, $Y_{pp,pA}$, to the yield for $pp$ collisions calculated without the medium effects, $Y_{pp}^{0}$. However, $Y_{pp}^{0}$ is unobservable quantity. For this reason, it is convenient to characterize the medium effects in $pc$ collisions in terms of the UE multiplicity dependence of the ratio of the experimental yields $Y_{pc}$ and the average yield for $pp$ collisions $\langle Y_{pp}\rangle$

$$\frac{Y_{pc}(\{p_{T}\},\{y\})}{\langle Y_{pp}(\{p_{T}\},\{y\})\rangle}\,,$$ (1)

where $\{p_{T}\}=(p_{T}^{a},p_{T}^{t})$ and $\{y\}=(y^{a},y^{t})$ are the sets of the transverse momenta and rapidities of the trigger particle and the associated hadron, and $\langle...\rangle$ means averaging over the UE multiplicity. In terms of the modification factors $I_{pc}$ (defined via the unobservable yield $Y_{pp}^{0}$) the ratio (1) should be equal to the ratio $I_{pc}/\langle I_{pp}\rangle$. Recently, the ALICE collaboration ALICE_Ipp ; ALICE_Ipp_PLB measured the UE multiplicity dependence of the ratio (1) for the hadron tagged jets in $pp$ and $p$+Pb collisions at $\sqrt{s}=5.02$ TeV. The ALICE ALICE_Ipp_PLB measurement has been performed for the hadron momenta $8<p_{T}^{t}<15$ GeV, $4<p_{T}^{a}<6$ GeV, and the UE activity has been characterized by the charged multiplicity $N_{ch}^{T}$ in the transverse kinematical region $\pi/3\leq|\phi|\leq 2\pi/3$, $|\eta|<0.8$, and $p_{T}>0.5$ GeV. As compared to the UE charged multiplicity density $dN_{ch}^{ue}/d\eta$, defined in the whole $\phi$ and $p_{T}$ regions for the pseudorapidity window $|\eta|<0.5$, $N_{ch}^{T}$ of ALICE_Ipp ; ALICE_Ipp_PLB is smaller by a factor of $\approx 4.4$. For $pp$ collisions, in Z_Ipp it was found that $I_{pp}$ decreases by about 7-10% with increase of the UE activity in the range $5\mathrel{\hbox to0.0pt{\lower 4.0pt\hbox{\hskip 1.0pt$\sim$}\hss}\raise 1.0pt\hbox{$<$}}dN_{ch}/d\eta\mathrel{\hbox to0.0pt{\lower 4.0pt\hbox{\hskip 1.0pt$\sim$}\hss}\raise 1.0pt\hbox{$<$}}20$ for the jet quenching calculated within the light-cone path integral formalism LCPI1 ; LCPI2004 ; RAA04 ; RAA08 for the induced gluon emission. The results of ALICE_Ipp for $I_{pp}/\langle I_{pp}\rangle$ agree qualitatively with calculations of Z_Ipp . It would be interesting to perform calculation of $I_{pA}$ and comparison with data from ALICE_Ipp_PLB as well. The data of ALICE_Ipp_PLB for $Y_{pA}$ also show a tendency of some decrease of $Y_{pA}$ with increasing the UE charged multiplicity. But the effect seems to be somewhat smaller, at least visually, than that observed for $pp$ collisions. However, one should bear in mind that in the case of $pA$ collisions the observed UE charged multiplicity density may contain a considerable fraction of hadrons that come from interaction of the projectile with peripheral nucleons without the formation of the collective QCD matter. Because interaction with the peripheral nucleons may produce low density/entropy parton system, for which the energy/entropy density is not large enough to form the QGP. Thus, one can expect that the fireball of the QCD matter in $pA$ collisions should have the core-corona structure (discussed previously for $AA$ collisions corona1 ). The effect of hadrons from the corona region on jet quenching should be small since these hadrons should be in the free streaming regime. For this reason, for the jet quenching calculation of the variation of $I_{pA}$ with the observed UE charged multiplicity $dN_{ch}^{ue}/d\eta$ one needs a formalism for accounting for the difference between the observed $dN_{ch}^{ue}/d\eta$ and the charged multiplicity related to formation of the QGP fireball (which we denote by $dN_{ch}^{f}/d\eta$). In the present paper we perform such jet quenching calculations of $I_{pA}$ for conditions of the ALICE experiment ALICE_Ipp using the Monte-Carlo (MC) Glauber model for calculation of the QGP fireball parameters as a function of the total UE charged multiplicity density. The parameters of the QGP fireball depends on the free parameters of the MC Glauber model. However, our results demonstrate that predictions for $I_{pA}$ turn out to be quite stable to the theoretical uncertainties of the MC Glauber scheme.

Theoretical framework for calculation of $I_{pA}$. Similarly to $AA$ collisions Wang_di-h ; PHENIX_di-h , the per-trigger yield $Y_{pA}$ for production of the trigger hadron $h^{t}$ and the associated hadron $h^{a}$ can be written via the di-hadron (back-to-back) and one-hadron inclusive $pN$ cross sections as

$$Y_{pA}(\{p_{T}\},\{y\})=\left\langle\frac{d^{4}\sigma_{pN}}{dp_{T}^{a}dp_{T}^{t}dy^{a}dy^{t}}\Big{/}\frac{d^{2}\sigma_{pN}}{dp_{T}^{t}dy^{t}}\right\rangle\,,$$ (2)

where $\langle\dots\rangle$ means averaging over the geometry of hard $pA$ collision. In particular, this averaging includes averaging over the impact parameter for $pA$ collision. However, it is natural to assume that the medium effects for $pA$ jet production are sensitive only to the parameters of the QGP fireball, and not to the impact parameter itself. For this reason, for a given UE charged multiplicity $N_{ch}^{ue}$ (i.e., for a given parameters of the QGP fireball) in (2) $\langle\dots\rangle$ is reduced to averaging over the geometry of the jet production point and of the jet trajectories in the QGP fireball. As in our previous calculations of $R_{pp}$ Z_hl and $I_{pp}$ Z_Ipp , we calculate $I_{pA}$ in the model of an effective azimuthally symmetric QGP fireball. We evaluate the distribution in the transverse plane of the jet production points in $pN$ collisions for Gaussian quark density (assuming that gluons have the same transverse distribution as quarks).

We perform calculations of the hard medium modified $pN$ cross sections in the collinear approximation, neglecting the internal parton transverse momenta in the colliding nucleons. The di-hadron and one-hadron cross sections can be written as

$$\frac{d^{4}\sigma_{pN}}{dp_{T}^{a}dp_{T}^{t}dy^{a}dy^{t}}=\int\frac{dz^{t}}{z^{t}}D_{h^{t}/i}^{m}(z^{t},p_{Ti})$$

$$\times D_{h^{a}/j}^{m}(z^{a},p_{Tj})\frac{d^{3}\sigma_{ij}}{p_{Ti}dp_{Ti}dy_{i}dy_{j}}\,,$$ (3)

$$\frac{d^{2}\sigma_{pN}}{dp_{T}^{t}dy^{t}}=\int\frac{dz^{t}}{z^{t}}D_{h^{t}/i}^{m}(z^{t},p_{Ti})\frac{d^{2}\sigma_{i}}{dp_{Ti}dy_{i}}\,,$$ (4)

where $\frac{d^{3}\sigma_{ij}}{dp_{Ti}dy_{i}dy_{j}}$ is the two-parton cross section of $p+N\to i+j+X$ process for $y_{i}=y^{t}$, $y_{j}=y^{a}$, $p_{Ti}=p_{Tj}=p_{T}^{t}/z^{t}$ and $z^{a}=z^{t}p_{T}^{a}/p_{T}^{t}$, $\frac{d^{2}\sigma_{i}}{dp_{Ti}dy_{i}}$ is the one-parton cross section for $p+N\to i+X$ process, $D_{h^{t}/i}^{m}$ and $D_{h^{a}/j}^{m}$ are the medium-modified FFs for transitions $i\to h^{t}$ and $j\to h^{a}$.

For calculation of the medium-modified FFs $D_{h/i}^{m}(z,Q)$ we use the same method as in Z_hl , and we therefore only briefly outline it. We write $D_{h/i}^{m}$ as a $z$-convolution

$$D_{h/i}^{m}(Q)\approx D_{h/j}(Q_{0})\otimes D_{j/k}^{in}\otimes D_{k/i}^{DGLAP}(Q)\,,$$ (5)

where $D_{k/i}^{DGLAP}$ is the DGLAP FF for $i\to k$ transition, $D_{j/k}^{in}$ is the in-medium $j\to k$ FF, and $D_{h^{a,t}/j}$ are the FFs for hadronization transitions of the parton $j$ to hadrons $h^{a,t}$. To compute the DGLAP FFs $D_{k/i}^{DGLAP}$ we use the PYTHIA event generator PYTHIA . As in Z_hl , we calculate the induced gluon spectrum using the method of RAA04 ; RAA08 , and express the in-medium FFs $D_{j/k}^{in}$ through the induced gluon spectrum in the approximation of the independent gluon emission RAA_BDMS . We treat the collisional energy loss, that is relatively small Z_coll , as a perturbation to the radiative mechanism (see Z_hl for details). For the FFs $D_{h/j}$ we use the KKP KKP parametrization with $Q_{0}=2$ GeV.

We calculate the gluon induced spectrum (needed for calculation of $D_{j/k}^{in}$) using the parametrization of running $\alpha_{s}$ in the form RAA20T ; Z_hl (supported by the lattice results for the in-medium $\alpha_{s}$ Bazavov_al1 )

$$\alpha_{s}(Q,T)=\begin{cases}\dfrac{4\pi}{9\log(\frac{Q^{2}}{\Lambda_{QCD}^{2}})}&\mbox{if }Q>Q_{fr}(T)\;,\\ \alpha_{s}^{fr}(T)&\mbox{if }Q_{fr}(T)\geq Q\geq cQ_{fr}(T)\;,\\ \frac{Q\alpha_{s}^{fr}(T)}{cQ_{fr}(T)}&\mbox{if }Q<cQ_{fr}(T)\;,\\ \end{cases}$$ (6)

with $c=0.8$, $Q_{fr}(T)=\Lambda_{QCD}\exp\left\{{2\pi}/{9\alpha_{s}^{fr}(T)}\right\}$ (we take $\Lambda_{QCD}=200$ MeV) and $Q_{fr}=\kappa T$, where $\kappa$ is a free parameter. We take $\kappa=2.55$. It has been obtained for scenario with the QGP formation in $pp$ collisions by $\chi^{2}$ fitting of the LHC data on $R_{AA}$ for 2.76 and 5.02 TeV Pb+Pb, and 5.44 TeV Xe+Xe collisions (see Z_hl for details).

As in Z_hl , we calculate $D_{j/k}^{in}$ for a fireball with a uniform entropy/density distribution in the transverse plane. We assume the Bjorken 1+1D longitudinal Bjorken expansion of the QGP fireball at the proper time $\tau>\tau_{0}=0.5$ fm. It corresponds to the QGP entropy density $s=s_{0}(\tau_{0}/\tau)$. As in Z_hl , for $\tau<\tau_{0}$ we take $s=s_{0}(\tau/\tau_{0})$. We write $s_{0}$ in terms of the charged multiplicity density $dN_{ch}^{f}/d\eta$ as

$$s_{0}=\frac{C}{\tau_{0}\pi R_{f}^{2}}\frac{dN_{ch}^{f}}{d\eta}\,,$$ (7)

where $R_{f}$ is the fireball radius, $C=dS/dy{\Big{/}}dN_{ch}/d\eta\approx 7.67$ is the entropy/multiplicity ratio BM-entropy , $dN_{ch}^{f}/d\eta$ is the part of the observed UE charged multiplicity density related to hadronization of the QGP fireball ($dN_{ch}^{f}/d\eta=F_{qgp}dN_{ch}^{ue}/d\eta$).

Calculation of the QGP fireball parameters. We calculate the fireball radius $R_{f}$ and the QGP fraction of the UE multiplicity $F_{qgp}$ as functions of the observed UE charged multiplicity density $dN_{ch}^{ue}/d\eta$ with the help of the MC Glauber wounded nucleon model developed in MCGL1 ; MCGL2 . The MC Glauber model of MCGL1 ; MCGL2 allows to perform calculations accounting for the nucleon meson cloud. In the present analysis we use the version without the meson cloud. The impact parameter profile of the probability of inelastic $NN$ interactions is taken in a Gaussian form $P(\rho)=\exp\left(-\pi\rho^{2}/\sigma_{in}^{pp}\right)\,.$ As in Z_xe , for $\sigma_{in}^{NN}$ we use the NSD inelastic $pp$ cross section $\sigma_{NSD}^{pp}=55.44$ mb. For the ²⁰⁸Pb nucleus we use the Woods-Saxon nuclear distribution $\rho_{A}(r)=\rho_{0}/[1+\exp((r-R_{A})/a)]$ with the hard-core $NN$ repulsion. We take $R_{A}=6.407$ fm, $a=0.459$ fm, and the hard-core radius $d=0.9$ fm GLISS3 . In MCGL1 ; MCGL2 ; Z_xe the MC Glauber simulations of nuclear collisions have been performed including the the sources corresponding to the wounded nucleons and to the binary collisions KN . Fitting of the LHC data on the centrality dependence of the midrapidity $dN_{ch}(AA)/d\eta$ gives the binary collisions fraction parameter $\alpha\sim 0.13$ MCGL2 ; Z_xe . However, the data on the midrapidity charged multiplicity density in the minimum bias 5.02 TeV $p$+Pb collisions ALICE_pPb agree better with $\alpha\approx 0$. In this case, the prediction of the Glauber wounded model for the minimum bias midrapidity charged multiplicity in $pA$ collisions reads

$$\frac{dN_{ch}^{mb}(pA)}{d\eta}=\frac{N_{w}^{mb}}{2}\cdot\frac{dN_{ch}^{mb}(pp)}{d\eta}\,,$$ (8)

where $N_{w}^{mb}$ is the minimum bias number of wounded nucleons, and $dN_{ch}^{mb}(pp)/d\eta$ is the minimum bias midrapidity charged multiplicity density in $pp$ collisions. The MC Glauber simulation of minimum bias 5.02 TeV $p$+Pb collisions gives $N_{w}^{mb}\approx 6.87$. With this value and $dN_{ch}^{mb}(pp)/d\eta\approx 5.34$, formula (8) gives $dN_{ch}^{mb}(p\mbox{Pb})/d\eta\approx 18.3$ that agrees well with the ALICE measurement $dN_{ch}^{mb}(pPb)/d\eta\approx 17.8$ ALICE_pPb . To generalize (8) to jet events one should account for the fact that one of the wounded nucleons in the nucleus and the projectile proton participate in jet production. It is reasonable to expect that the contribution of this pair $pN$ to the multiplicity density should be equal to the average UE multiplicity density $dN_{ch}^{ue}(pp)/d\eta$ in $pp$ jet events. Then, the generalization of (8) to the case of the minimum bias UE multiplicity density in jet events can be written as

$$\frac{dN_{ch}^{ue}(pA)}{d\eta}=\frac{dN_{ch}^{ue}(pp)}{d\eta}+\frac{N_{w}^{ue}-2}{2}\cdot\frac{dN_{ch}^{mb}(pp)}{d\eta}\,,$$ (9)

where $N_{w}^{ue}$ is the average number of the wounded nucleons in jet events. Our MC Glauber simulation gives $N_{w}^{ue}\approx 9.9$ for 5.02 TeV $p$+Pb collisions. Then, using $dN_{ch}^{ue}(pp)/d\eta\approx 12.5$ for $5.02$ TeV $pp$ jet events (obtained by interpolating the ATLAS data ATLAS_UE_Nch at $\sqrt{s}=0.9$ and $7$ TeV, assuming that $dN_{ch}^{ue}(pp)/d\eta\propto s^{\delta}$), from the formula (9) we obtain $dN_{ch}^{ue}(p\mbox{Pb})/d\eta\approx 33.5$, which agrees well with that obtained in the ALICE experiment ALICE_Ipp ($\langle N_{ch}^{T}\rangle=7.77\pm 0.31$ that corresponds to $dN_{ch}^{ue}(p\mbox{Pb})/d\eta=34.19\pm 1.36$). Thus, one sees that the predictions of the MC Glauber model for the minimum bias and the UE midrapidity charged multiplicities in $p$+Pb collisions are in good agreement with the experimental data.

To perform the MC Glauber simulation of the geometry of the entropy production in $pA$ jet events, for each wounded nucleon we use the Gaussian distribution of the entropy density in the transverse coordinates

$$\frac{dS_{w}^{i}(\mbox{{\boldmath$\rho$}}-\mbox{{\boldmath$\rho$}}_{i})}{dyd\mbox{{\boldmath$\rho$}}}=\frac{C\exp[-(\mbox{{\boldmath$\rho$}}-\mbox{{\boldmath$\rho$}}_{i})^{2}/\sigma^{2}]}{\pi\sigma^{2}}\cdot\frac{dN^{i}_{ch}/d\eta}{2}\,,$$ (10)

where $i=1,2$ correspond to the nucleons participating in jet production, and $i>2$ to the ordinary wounded nucleons. We treat $dN^{i}_{ch}/d\eta$ as random variables. Below we use the short-hand notations $n_{i}$ for $dN^{i}_{ch}/d\eta$. As in MCGL2 ; Z_xe , we describe fluctuations of $n_{i}$ by the widely used in the MC Glauber simulations Gamma distribution

$$G(n_{i})=\left(\frac{n_{i}}{\langle n_{i}\rangle}\right)^{\kappa_{i}-1}\frac{\kappa_{i}^{\kappa_{i}}\exp\left[-n_{i}\kappa_{i}/\langle n_{i}\rangle\right]}{\langle n_{i}\rangle\Gamma(\kappa_{i})}\,.$$ (11)

The $\langle n_{i>2}\rangle$ should be equal to the minimum bias $pp$ charged multiplicity pseudorapidity density $dN_{ch}^{mb}(pp)/d\eta$, i.e., for $5.02$ TeV $pp$ collisions we have $\langle n_{i>2}\rangle\approx 5.34$. The value of the parameter $\kappa_{i>2}$ has been adjusted to reproduce the experimental variance of the charged $pp$ multiplicity in the unit pseudorapidity window $|\eta|<0.5$. This gives $\kappa_{i>2}\approx 0.56$ Z_xe . The value of $\langle n_{i=1,2}\rangle$ should be equal to $dN_{ch}^{ue}(pp)/d\eta\approx 12.5$. We adjusted the parameter $\kappa_{i=1,2}$ to reproduce the variance of the midrapidity $pp$ UE charged multiplicity density, that has been evaluated using the CMS data CMS_UE7 on the UE multiplicity distribution for $7$ TeV $pp$ collisions (rescaled to $\sqrt{s}=5.02$ TeV assuming the KNO scaling). This procedure gives $\kappa_{i=1,2}\approx 1.1$

We used the following procedure to differentiate the QGP fireball and corona transverse regions in our MC Glauber simulation of the UE for jet production in $p$+Pb collisions. We attribute to the QGP fireball the regions with the local ideal gas temperature (defined via the entropy density at $\tau=\tau_{0}$) larger than $T_{min}\sim T_{c}$ (where $T_{c}\approx 160$ MeV is the deconfinement temperature). We performed calculations for two versions with $T_{min}=160$ and $200$ MeV. Using this procedure we obtained the relative QGP contribution to the UE charged multiplicity $dN_{ch}^{ue}/d\eta$ as functions of the observed UE charged multiplicity density $dN_{ch}^{ue}/d\eta$. We define the radius of the effective fireball (with the flat entropy density), from the condition that the mean squared radius of the QGP fireball coincides with that for the MC Glauber entropy distribution in the region $T>T_{min}$. We also performed the calculations for the radius of the effective fireball defined as $R_{f}=\sqrt{S_{f}/\pi}$, where $S_{f}$ is the area of the region with $T>T_{min}$. It was found that the results for these two methods are very close to each other.

Figure 1: The parameters of the QGP fireball obtained in the MC Glauber model simulations for $\sigma=0.7$ (thick lines) and $0.4$ fm (thin lines) for $T_{min}=160$ (solid) and $200$ (dashed) MeV. (a) The relative contribution of the QGP fireball to the midrapidity UE charged multiplicity density $dN_{ch}^{ue}/d\eta$ vs $dN_{ch}^{ue}/d\eta$. (b) The QGP fireball radius vs $dN_{ch}^{f}/d\eta$.

Figure 2: The ratio $I_{pA}/\langle I_{pp}\rangle$ vs the midrapidity UE charged multiplicity density $dN_{ch}^{ue}/d\eta$ in $5.02$ $p$+Pb collisions obtained for $\sigma=0.7$ (a) and $0.4$ (b) fm for $T_{min}=160$ (solid) and $200$ (dashed) MeV. Data points are from ALICE_Ipp_PLB (with the rescaled $N_{ch}^{T}$ variable used in ALICE_Ipp_PLB by the factor 4.4 (see text for explanation).

Numerical results. To illustrate the sensitivity of our results to the choice of the Gaussian width parameter $\sigma$ (entering Eq. (10)) in the MC Glauber simulations of the UE for jet production in $pA$ collisions, we present the results for $\sigma=0.7$ and $0.4$ fm (the latter value is often used in the MC Glauber model GLISSANDO GLISS3 ). In Fig. 1a we plot the average relative QGP contribution to the UE charged multiplicity density $F_{qgp}$ as a function of the UE charged midrapidity multiplicity density $dN_{ch}^{ue}(pA)/d\eta$ obtained in the MC Glauber simulations for $T_{min}=160$ and $200$ MeV. As seen in Fig. 1a, $F_{qgp}$ becomes smaller at lower total UE charged multiplicity, i.e., the corona effect is more important in the low multiplicity jet events. In Fig. 1b we present the results for the effective fireball radius as a function of $dN_{ch}^{f}/d\eta=F_{qgp}dN_{ch}^{ue}/d\eta$ for the same parameters as in Fig. 1a. For the initial QGP temperature (for the ideal gas model) we obtained $T_{0}\sim 210-220(330-350)$ MeV at $dN_{ch}^{f}/d\eta\sim 10(60)$.

The results for $I_{pA}/\langle I_{pp}\rangle$ as a function of the UE charged multiplicity density are shown in Fig. 2 (we use $\langle I_{pp}\rangle\approx 0.94$ obtained in Z_Ipp ). The coordinates for the data points from ALICE_Ipp_PLB , shown in Fig. 2, have been rescaled by the factor $r\approx 4.4$ from the variable $N_{ch}^{T}$ (corresponding to the UE charged multiplicity in the transverse kinematical region $\pi/3\leq|\phi|\leq 2\pi/3$, $|\eta|<0.8$, and $p_{T}>0.5$ GeV) to the UE charged multiplicity $dN_{ch}^{ue}/d\eta$ defined in the whole $\phi$ and $p_{T}$ regions for the pseudorapidity window $|\eta|<0.5$. The theoretical curves were obtained for $\sigma=0.7$ (a) and $0.4$ (b) fm and $T_{f}=160$ (solid) and $200$ (dashed) MeV. Fig. 2 clearly shows that predictions for $I_{pA}$ turn out to be rather insensitive to the parameter $\sigma$ in (10) used in the MC Glauber model and to the value of $T_{min}$ used for evaluation of the size and the entropy of the QGP fireball. We also performed calculations for the MC Glauber scheme without the short range $NN$ correlations. We have found that in this case the results are practically the same as that shown in Figs. 1, 2. It is not surprising, since for $p$+Pb collisions the typical 3D separation between the wounded nucleons is larger than the $NN$ hard core radius by a factor of $2-3$.

As one can see from Fig. 2, the theoretical $I_{pA}$ decreases by $\sim 8$% from $dN_{ch}^{ue}/d\eta\sim 10$ to $dN_{ch}^{ue}/d\eta\sim 60$. The experimental data show approximately similar tendency. The theoretical predictions are in reasonable agreement with the data within the measured uncertainties. However, from Fig. 2 one sees that the normalization of the theoretical curves is smaller by $\sim 3-4$% than that for the ALICE data ALICE_Ipp_PLB . This may be due to the systematic errors of the experimental yield $Y_{pA}$, that are rather large ($\sim 7$%). Physically, in the picture with jet modification due to the final state interaction effects in the QGP fireball, one can expect that in the region with $dN_{ch}^{ue}(pA)/d\eta\sim\langle dN_{ch}^{ue}(pp)/d\eta\rangle\mathrel{\hbox to0.0pt{\lower 4.0pt\hbox{\hskip 1.0pt$\sim$}\hss}\raise 1.0pt\hbox{$<$}}12-13$ (i.e., in the regime where the single $pN$ interaction should dominate) there must be $Y_{pA}/\langle Y_{pp}\rangle\approx 1$, i.e., $I_{pA}/\langle I_{pp}\rangle\approx 1$. But for the ALICE data ALICE_Ipp_PLB we have $Y_{pA}/\langle Y_{pp}\rangle\sim 1.05\pm 0.07$ at $dN_{ch}^{ue}(pA)/d\eta\sim\langle dN_{ch}^{ue}(pp)/d\eta\rangle$. In principle, the double-parton scattering can lead to $Y_{pA}/\langle Y_{pp}\rangle>1$. However, the procedure of ALICE_Ipp_PLB with calculating the ratio $Y_{pA}/\langle Y_{pp}\rangle$ using for the effective away side $pA$ and $pp$ yields the difference between the real measured away side yields and the measured yields in the transverse region, should automatically eliminate the possible background contribution from the double-parton scattering. It is worth noting that for the data of ALICE_Ipp_PLB the factor $I_{pPb}$ at the maximal value of the $p$+Pb UE multiplicity ($N_{ch\text{max}}^{T}(pPb)\approx 13.4$) is larger than the factor $I_{PbPb}$ at the minimal Pb+Pb UE multiplicity ($N_{ch\text{min}}^{T}(PbPb)\approx 19.2$) by $\sim 10$%, while, physically, one could expect $I_{pPb}/I_{PbPb}\approx 1$ for $N_{ch}^{T}(pPb)\sim N_{ch}^{T}(PbPb)$ (since the UE multiplicity dependence of $I_{pPb}$ is weak, it is clear that the difference between $N_{ch\text{max}}^{T}(pPb)$ and $N_{ch\text{min}}^{T}(PbPb))$ cannot lead to a considerable variation of $I_{pPb}/I_{PbPb}$). Note that in our model the factor $I_{pPb}$ at $dN_{ch}^{ue}(pPb)/d\eta\sim 85$ (that corresponds to $N_{ch\text{min}}^{T}(PbPb)\approx 19.2$) agrees reasonably with the factor $I_{PbPb}$ of ALICE_Ipp_PLB at the minimal $N_{ch}^{T}(PbPb)$.

Summary. We have calculated the medium modification factor $I_{pA}$ for hadron-tagged jets in 5.02 TeV $p$+Pb collisions. The medium-modified FFs have been evaluated within the light-cone path integral approach to induced gluon emission. We used parametrization of the running QCD coupling $\alpha_{s}(Q,T)$ which has a plateau around $Q\sim\kappa T$ (motivated by the lattice simulations Bazavov_al1 ). The value of $\kappa$ is fitted to the LHC data on the nuclear modification factor $R_{AA}$ in $2.76$ and $5.02$ TeV Pb+Pb, and $5.44$ TeV Xe+Xe collisions. To determine the size of the QGP fireball and the fraction of the UE charged multiplicity density related to the formation of the QGP fireball in $pA$ collisions we used the MC Glauber model. We have found that the theoretical predictions for $I_{pA}$ are rather insensitivity to the theoretical uncertainties in the MC Glauber modeling of the QGP formation. Our calculations show that the jet quenching effect for $I_{pA}$ turns out to be rather small. We have found that the theoretical $I_{pA}$ as a function of the UE charged multiplicity density, within errors, agrees with data from ALICE ALICE_Ipp_PLB for 5.02 TeV $p$+Pb collisions. However, the experimental errors are too large to draw a firm conclusion on the possible presence of jet quenching.