betaFeedback

WorksPhysics and Astronomy

High-energy dipole scattering amplitude from evolution of low-energy proton light-cone wave functions

Abstract

The forward scattering amplitude of a small dipole at high energies is given in the mean field approximation by the Balitsky-Kovchegov (BK) evolution equation. It requires an initial condition N(r;x0)N(r; x_0) describing the scattering of a dipole with size rr off the target that is probed at momentum fraction x0x_0. Rather than using ad hoc parameterizations tuned to high-energy data at x≪x0x\ll x_0, here we attempt to construct an initial scattering amplitude that is consistent with low-energy, large-xx properties of the proton. We start from a non-perturbative three quark light-cone model wave function from the literature. We add O(g){\cal O}(g) corrections due to the emission of a gluon, and O(g2){\cal O}(g^2) virtual corrections due to the exchange of a gluon, computed in light-cone perturbation theory with exact kinematics. We provide numerical data as well as analytic parameterizations of the resulting N(r;x0)N(r; x_0) for x0=0.01−0.05x_0=0.01 - 0.05. Solving the BK equation in the leading logarithmic (LL) approximation towards lower xx, we obtain a fair description of the charm cross section in deeply inelastic scattering measured at HERA by fitting one parameter, the coupling constant αs≃0.2α_s\simeq 0.2. However, without the option to tune the initial amplitude at x0x_0, the fit of the high precision data results in χ2/Ndof=2.3χ^2/N_\text{dof} = 2.3 at Ndof=38N_\text{dof} =38, providing clear statistical evidence for the need of systematic improvement e.g. of the photon wave function, evolution equation, and initial condition.

Introduction

In Deep Inelastic Scattering (DIS) a pointlike virtual photon probes the rich QCD dynamics taking place inside the proton or a nucleus. At high energies, where the small Bjorken-xx part of the target wave function is probed, one observes very large gluon densities [1]. When the gluon densities become of the same order as inverse coupling, non-linear QCD dynamics start to dominate and multiple scattering effects are important [2]. In the high-energy limit, the scattering process is most conveniently described in the dipole picture in a frame where the virtual photon has a large momentum [3], and its partonic Fock states, such as ∣qqˉ⟩|q\bar{q}\rangle at leading order (LO), have a long lifetime as they scatter from the color field of the target.

Describing the QCD dynamics in this high-density domain is natural in the Color Glass Condensate [4] framework. Here the center-of-mass energy or Bjorken-xx dependence of various observables (and as such the target structure) is described in the large-NcN_c limit by the perturbative Balitsky-Kovchegov (BK) renormalization group equation [5, 6]. It describes how the dipole-target scattering amplitude, which contains information about the target structure, changes with increasing energy. The dipole amplitude (a correlator of two Wilson lines) is actually a convenient degree of freedom at high energies: all cross sections computed at high energy in the CGC framework are expressed in terms of the dipole amplitude or higher-point correlators which can be written, in a Gaussian approximation, in terms of the two-point dipole amplitude [7].

The initial condition for the dipole-proton scattering amplitude depends on non-perturbative properties of the proton. A typical approach in the field has been to assume an intuitive functional form as an initial x0≪1x_0 \ll1 and fit various unknown parameters to the HERA total cross section data; see, e.g., Refs. [8, 9, 10] where a very good description of small-xx HERA data is obtained at leading order, resumming powers of αsln⁡1/x\alpha_s \ln1/x via BK evolution with running coupling corrections [11]. Recent developments to full NLO accuracy have also allowed for a simultaneous description of total and heavy quark production data [12, 13]. The drawback of this approach is that one is sensitive to the assumed functional form of the initial dipole amplitude and that the model parameters need to be re-fitted if the evolution is initialized at different x0x_0. Furthermore, there is no relation to the low energy (or “large-xx”) proton structure.

In this work, we take a complementary approach aiming to compute the initial dipole-proton scattering amplitude at moderate x0x_0. As we will discuss in more detail next, the necessary non-perturbative input consists in a proton valence quark wave function that is constrained by low-energy data. The x0x_0-dependent initial condition is then obtained by computing the dipole-target scattering amplitude including one perturbative gluon emission in the target, with the gluon longitudinal momentum fraction regulated by x0x_0 [14, 15]. The advantages of this approach are that we do not assume an ad-hoc functional form of the scattering amplitude and that the initial condition can be computed and the BK evolution initialized at any (moderate) x0x_0 without a need to perform new fits. Also, this approach largely eliminates the freedom of tuning initial conditions in order to optimally match the evolution equation to the small-xx data. This may reveal quantitative evidence for the need for improvements beyond leading-log, or even running coupling BK evolution.

Finally, we would like to point out that light-cone Hamiltonian calculations of wave functions have been employed previously to set initial conditions for QCD scale evolution to high virtuality Q2Q^2, in order to describe DIS in the ”dipole approach” using correlators of eikonal Wilson lines as degrees of freedom [16, 17, 18, 19]. Our approach is similar in spirit although here the goal is to determine initial conditions for evolution to small xx.

Dipole-proton scattering at moderate x

We first provide an overview of our approach to the light-cone structure of the proton. We employ a truncated Fock space description which starts with a three quark state. The corresponding Fock space amplitude (wave function) Ψqqq\Psi_{qqq} corresponds to a non-perturbative solution of the QCD light-front Hamiltonian. To date, exact solutions for the light-cone wave functions are not available. In the future, lattice gauge theory may provide numerical solutions for moderate parton momentum fractions xix_i and transverse momenta k⃗i\vec{k}_i via a large momentum expansion of equal-time Euclidean correlation functions in instant quantization [20, 21]; see ref. [22] for a recent lattice computation of the wave function of the leading qqˉq\bar{q} state of the pion. Also, the MAP collaboration [23] has recently extracted the wave functions of the first four Fock states of the pion from fits to its parton distribution functions and electromagnetic form factor.

Here, we rely on solutions of effective light-cone Hamiltonians for guidance on the low energy and low virtuality Q2Q^2 structure of the proton. Specifically, we shall employ the HO wave function of Refs. [24, 25]. In these references, the authors fixed the parameters of the three quark wave function to the proton “radius,” or Dirac form factor at Q2→0Q^2 \to0, to the anomalous magnetic moments of the proton and neutron, and to the axial vector coupling gAg_A. The wave function also matches reasonably well the empirical knowledge of the longitudinal and transverse momentum distribution of single quarks in the valence quark regime. Finally, the wave function of Refs. [24, 25] also provide predictions for quark momentum correlations.

At next-to-leading order (NLO) in the Fock expansion we add the three quarks and one gluon state with amplitude Ψqqqg\Psi_{qqqg}, as well as the virtual corrections to Ψqqq\Psi_{qqq} due to the exchange of a gluon by two quarks in the proton. These corrections are obtained via light-cone perturbation theory calculations [14, 15]. The presence or exchange of the gluon extends the range of parton light-cone momentum fractions to lower xx, and pushes their transverse momenta into the perturbative regime. It also affects their momentum correlations.

The central element of our analysis is the (imaginary part of the) eikonal scattering amplitude NN of a small dipole of transverse size r\mathbf{r}. The real part of NN corresponds to two-gluon exchange,

N(r,b)=−g4CF∫d2K d2q(2π)4cos⁡(b⋅K)(q−12K)2(q+12K)2(cos⁡(r⋅q)−cos⁡(r⋅K2))G2(q−12K,−q−12K).(1)N(\mathbf{r},\mathbf{b})=-g^4 C_F \int\frac{\mathrm{d}^2\mathbf{K}\,\mathrm{d}^2\mathbf{q}}{(2\pi)^4}\frac{\cos(\mathbf{b}\cdot\mathbf{K})}{\left(\mathbf{q}-\frac{1}{2}\mathbf{K}\right)^2\left(\mathbf{q}+\frac{1}{2}\mathbf{K}\right)^2} \left(\cos(\mathbf{r}\cdot\mathbf{q})-\cos\left(\frac{\mathbf{r}\cdot\mathbf{K}}{2}\right)\right)G_2\left(\mathbf{q}-\frac{1}{2}\mathbf{K},-\mathbf{q}-\frac{1}{2}\mathbf{K}\right). \tag*{(1)}

Here K\mathbf{K} is the momentum transfer which is Fourier conjugate to the impact parameter b\mathbf{b}. As explained below, we will eventually average N(r,b)N(\mathbf{r},\mathbf{b}) over a suitable range of impact parameters. We emphasize that the expression above accounts only for a single, perturbative two-gluon exchange (see its derivation in Ref. [26]), it does not resum the Glauber-Mueller multiple scattering series. This restricts its applicability to the regime of weak scattering, N(r,b)≪1N(\mathbf{r},\mathbf{b}) \ll1. Furthermore, N(r,b)N(\mathbf{r},\mathbf{b}) actually acquires an imaginary part due to the perturbative exchange of three gluons; its magnitude has been shown to be much smaller than its real part [27, 28], and in practice it is of interest only for processes involving C-parity odd exchanges [29]. For the present purposes, it can be neglected.

The coupling of the two static gluons to the proton is described in terms of the color charge density correlator

⟨ρa(q1)ρb(q2)⟩=δabg2G2(q1,q2).(2)\langle\rho^a(\mathbf{q}_1)\rho^b(\mathbf{q}_2)\rangle=\delta^{ab}g^2G_2(\mathbf{q}_1,\mathbf{q}_2). \tag*{(2)}

The color charge density operator corresponds to the light-cone plus component of the color current on the x+=0x^+=0 light front, integrated over x−x^-, ρa(q)≡J+a(q)\rho^a(\mathbf{q})\equiv J^{+a}(\mathbf{q}), when the proton carries positive PzP_z. Eqs. (1,2) correspond to the leading twist contribution to the matrix element of the dipole operator in the proton. Dozens of diagrams contribute to this correlator at NLO, their explicit expressions are listed in Ref. [14]. We point out that G2(q1,q2)G_2(\mathbf{q}_1,\mathbf{q}_2) satisfies a Ward identity due to the color neutrality of the proton; it vanishes when either q12q_1^2 or q22→0q_2^2 \to0 so that N(r,b)N(\mathbf{r},\mathbf{b}) in Eq. (1) is free of IR divergences. However, G2G_2 does exhibit a collinear singularity which is regularized by assigning a mass to the quarks in the light-cone energy denominators for the q→qgq \to qg and qg→qqg \to q vertices; see Ref. [14] for details. All the results presented here were obtained with mcoll=0.2m_{\mathrm{coll}} = 0.2 GeV. This is consistent with the quark mass and transverse momentum scales which appear in the non-perturbative three quark wave function of Refs. [24, 25]. The color charge correlator also exhibits a soft singularity when the light-cone momentum fraction xgx_g of the gluon goes to zero. This is regularized with a cutoff xx on xgx_g, and the resummation of yet softer gluons will be performed through the BK equation. Note that at x=0.1x = 0.1 the NLO contribution to G2G_2 truly is a reasonably small O(g2)O(g^2) perturbative correction [15]. However, by x=0.01x = 0.01 its magnitude grows to essentially O(1)O(1), a leading-log correction. Hence, at such xx resummation is required and it is justified to use the computed dipole as an initial condition for the leading order BK evolution.

We recall, also, that at the given order ultraviolet divergences cancel [14], so that G2G_2 is independent of the renormalization scale, and that the coupling does not run. Lastly, let us mention that the angular dependence of the correlator G2G_2, as well as the dependence of its Fourier transform on impact parameter, has been analyzed numerically in detail in Ref. [15].

Small-x evolution of the proton light-cone wave function

In order to obtain an initial condition for b-independent BK evolution4 we average the dipole-target scattering amplitude obtained from Eq. (1) over the impact parameter b\mathbf{b},

N(r,x0)=1ST∫bmax⁡d2b N(r,b,x0).(3)N(r,x_0) = \frac{1}{S_T}\int^{b_{\max}} \mathrm{d}^2\mathbf{b}\,N(r,\mathbf{b},x_0). \tag*{(3)}

Throughout this work, we denote the magnitudes of the transverse vectors as b=∣b∣b = |\mathbf{b}| and r=∣r∣r = |\mathbf{r}|. The resulting amplitude is dominated by perturbative contributions when the dipole size rr is small. In this region there is a small cos⁡(2ϕ)\cos(2\phi) dependence on the angle ϕ\phi between r\mathbf{r} and b\mathbf{b} [15] which vanishes when we integrate over b\mathbf{b}. Here STS_T is the proton transverse area. Inclusive cross sections considered in this work are not sensitive to the actual shape of the target but only to the total transverse size. The proton geometry is most directly probed in exclusive vector meson production process where the total momentum transfer K\mathbf{K} which is Fourier conjugate to the impact parameter is measurable. Parametrizing the J/ψJ/\psi production cross section in HERA kinematics as e−BDK2e^{-B_D\mathbf{K}^2} one obtains BD=4 GeV−2B_D = 4\,\mathrm{GeV}^{-2} [32]. Assuming a Gaussian impact parameter profile for the proton, this corresponds to a two-dimensional root-mean-square radius bGaussian=2B≈0.56b_{\mathrm{Gaussian}} = \sqrt{2B} \approx0.56 fm and a proton area ST=2πBS_T = 2\pi B. On the other hand, if we assume a step function (hard sphere) profile for the proton, the same diffractive slope is obtained when the proton radius is bHardsphere=2B≈0.79b_{\mathrm{Hardsphere}} = 2\sqrt{B} \approx0.79 fm, which corresponds to ST=4πBS_T = 4\pi B.

Although exclusive vector meson data favors the Gaussian density profile over the hard sphere one (see e.g. [33]), the current data does not constrain the proton shape precisely. We also note that if the b-dependent dipole amplitude from Eq. (1) is directly used to compute exclusive J/ψJ/\psi production cross section, the resulting spectra differs from the Gaussian profile case only in the region where there are no experimental constraints [34]. In this work the results shown below by default correspond to the Gaussian density profile (with bmax⁡=bGaussianb_{\max} = b_{\mathrm{Gaussian}}) unless otherwise stated, but we also study the dependence on the bmax⁡b_{\max} cut by using a step function profile with bmax⁡=bHard sphere=2bGaussianb_{\max} = b_{\mathrm{Hard\ sphere}} = \sqrt{2}b_{\mathrm{Gaussian}}.

The proton transverse area STS_T has also been extracted by fitting a parameterized initial condition for the BK evolution equation to the HERA structure function data. Leading order analyses [9, 10] typically obtain ST∼16S_T \sim16 mb. In recent fits at NLO accuracy [12, 13] proton areas ST∼10…20S_T \sim10\ldots20 mb were obtained depending on the details of the analysis setup. We test this uncertainty in the proton small-xx transverse profile by showing some results for both the Gaussian and hard sphere profiles with transverse areas 9.8 mb and 19.6 mb, respectively.

Before performing the impact parameter average we first study the impact parameter profile from the NLO light-cone wave function seen by a perturbative probe:

T(b)=C∫rmax⁡d2r N(r,b,x).(4)T(\mathbf{b}) = C\int^{r_{\max}} \mathrm{d}^2\mathbf{r}\,N(r,\mathbf{b},x). \tag*{(4)}

The normalization condition ∫bmax⁡d2b T(b)=1\int^{b_{\max}} \mathrm{d}^2\mathbf{b}\,T(\mathbf{b}) = 1 is used to fix the constant CC. We will refer to T(b)T(\mathbf{b}) as the transverse ”density” profile to match standard terminology from the literature. As the dipole amplitude is a rapidly increasing function of the dipole size rr, this integral is dominated by dipoles of size r∼rmax⁡r \sim r_{\max}, as long as rmax⁡r_{\max} is in the perturbative domain.

The extracted density profiles up to bmax⁡=0.8b_{\max} = 0.8 fm for different rmax⁡r_{\max} are shown in Figure 1. For reference, a Gaussian profile, as used e.g. in the popular IPsat parametrization [35] for the dipole amplitude with the slope B=4 GeV−2B = 4\,\mathrm{GeV}^{-2}, is also shown. We observe a similar transverse profile except for very central b≲0.2b \lesssim0.2 fm where the computed profile is more steeply falling. This region can only be probed at high momentum transfer ∣t∣≳1 GeV2|t| \gtrsim1\,\mathrm{GeV}^2 [34], which is not covered in the currently available coherent vector meson production data. The high-bb tails of T(b)T(\mathbf{b}) resulting from the LCPT one gluon emission corrections are exponential rather than Gaussian. However, in all we conclude that for the present purposes the Gaussian profile used to match STS_T to bmax⁡b_{\max} is a reasonable approximation.

Effective normalized proton density profiles

Figure 1. Effective normalized proton density profiles at x=0.01x = 0.01 extracted from N(r,b)N(\mathbf{r},\mathbf{b}) with bmax⁡=0.8b_{\max} = 0.8 fm up to NLO in the Fock expansion of the light-cone wave function.

The bb-averaged dipole amplitudes (using a Gaussian profile) are shown in Figure 2 (linear scale) and Figure 3 (logarithmic scale) at x0=0.05x_0 = 0.05, x0=0.025x_0 = 0.025 and x0=0.01x_0 = 0.01. Here we also show the dependence on the diffractive slope BB: the bands correspond to varying BB by ±10%\pm10\% which changes both STS_T and bmax⁡b_{\max}. The results depend weakly on this cut especially in the perturbative small-rr domain. The dipole amplitude increases with rr, approximately proportional to r2r^2, as expected. For r≳0.4r \gtrsim0.4 fm the color neutrality of the proton, and the fact that the dipole scatters from a target of finite transverse extent, begin to slow the growth of N(r)N(r); a model that does not account for the finite size of the proton in impact parameter space would attribute this to power corrections. Finally, when the size of the dipole becomes comparable to that of the target the amplitude is found to decrease again (not shown) as the end points of the dipole essentially “miss” the target. However, we emphasize that this behavior occurs at large r∼r \sim few fm where in any case the perturbative calculation of the scattering amplitude is not valid.

Figure 3. Same as Figure 2 but on a double logarithmic scale in order to better exhibit the behavior at small rr.

Impact-parameter averaged dipole as a function of dipole size

Figure 2. Impact-parameter averaged dipole as a function of dipole size rr at two different momentum fractions xx. The bands correspond to varying the proton shape parameter BB by 10%. The dotted lines show best fits to the central values with the modified MV model parametrization of Eq. (5).

Figures 2 and 3 confirm that down to x=0.01x = 0.01 scattering of small dipoles with rr significantly less than 1 fm remains quite weak, at least for αs=0.2\alpha_s = 0.2 which we determine below from a fit to the charm cross section in DIS. Therefore, it appears reasonable to start small-xx evolution with this initial condition at xx in the range 0.01…0.050.01 \ldots0.05.

To obtain analytic parameterizations of the dipole amplitude we fit our numerical data for the b\mathbf{b}-averaged scattering amplitude to the following expression which is inspired by the McLerran-Venugopalan (MV) model [36]:

N(r)=1−exp⁡[−r2Qs,024ln⁡(1rΛ+ec⋅e)](5)N(r) = 1 - \exp\left[-\frac{r^2 Q_{s,0}^2}{4}\ln\left(\frac{1}{r\Lambda}+e_c \cdot e\right)\right] \tag*{(5)}

where Λ=0.241\Lambda= 0.241 GeV is a fixed infrared scale. Such a parameterization has been used previously e.g. in Refs. [9, 10, 12] to fit the initial condition for BK evolution to the HERA data. While our fit is restricted to r<0.5r < 0.5 fm, the parameterization forces N(r)→1N(r) \to1 in the large-rr region. Of course, the behavior at large rr can not be trusted, and other extrapolations would be possible. It is important, however, that the large-rr extrapolation is such that the Fourier transform of 1−N(r)1 - N(r) at high kk (and, consequently, the forward particle production cross section, for example) will be insensitive to the assumed form. We will also demonstrate below that perturbative observables, in our case the charm production cross section, are only sensitive to the perturbative regime of small dipoles where our calculation should apply, and not to the extrapolation to large rr.

The free parameters in Eq. (5), Qs,02Q_{s,0}^{2}, γ\gamma and ece_c, are fit to the calculated dipole amplitude in the region 0.01 fm <r<< r < 0.5 fm (we actually fit the logarithm of the dipole in order to give equal weight to small and intermediate rr). The upper limit restricts to the perturbative domain, and the lower limit is imposed in order to give some weight to the region of intermediate rr as well. The resulting dipole amplitudes are shown in Figure 2 and Figure 3 as dotted lines. The fit parameters are listed in Table I. In the fit we require that ec>e−1e_c > e^{-1} in order to enforce positivity of the logarithm in Eq. (5), and all fit results give ec=e−1e_c=e^{-1} within numerical accuracy, i.e. they require as small an infrared cutoff as allowed. The MV-model inspired parameterization is found to describe the dipole-proton scattering amplitude quite well, for all dipole sizes in the perturbative r≲0.5r \lesssim0.5 fm region. Here, of course, the linearized version of Eq. (5) is sufficient, as it should be: recall that Eq. (1) does not resum multiple scattering.

xxQs,02Q_{s,0}^{2} [GeV2^{2}]γ\gammaece_c
0.010.100−0.004+0.0040.100^{+0.004}_{-0.004}1.001−0.001+0.0011.001^{+0.001}_{-0.001}e−1e^{-1}
0.0250.066−0.003+0.0030.066^{+0.003}_{-0.003}0.998−0.001+0.0010.998^{+0.001}_{-0.001}e−1e^{-1}
0.050.047−0.002+0.0020.047^{+0.002}_{-0.002}0.997−0.001+0.0010.997^{+0.001}_{-0.001}e−1e^{-1}

Table I. Best fit parameters to the modified MV model parameterization, Eq. (5), for a fit over 0.01 fm <r<< r < 0.5 fm. The upper and lower limits are obtained by varying the proton shape parameter BB by ±10%\pm10\%. All fit results give ec=e−1e_c=e^{-1} within numerical accuracy.

The momentum scale Qs,02Q_{s,0}^{2} remains non-perturbative down to x=0.01x=0.01; see below for an extraction of a “saturation scale” at lower xx. However, it increases approximately as Qs,02∼1/x0.47Q_{s,0}^{2}\sim1/x^{0.47}. The “anomalous dimension” of the dipole amplitude is γ=1\gamma=1 within numerical accuracy. It appears reasonable to us that the initial condition for the evolution equation admits a power series expansion in r2r^{2}, starting at its first power. On the other hand, leading order fits to HERA total cross section data [9, 10] require γ∼1.1−1.2\gamma\sim1.1-1.2 in order to obtain as slow a Q2Q^{2} dependence of the cross section as required by the HERA data [37] [1]; recent fits at next-to-leading order accuracy performed in Ref. [12, 13] also prefer γ≳1\gamma\gtrsim1 when the heavy quark production data is included. A problem with γ>1\gamma>1 is that it renders the (dipole) unintegrated gluon distribution function [7, 38, 39] and the forward particle production cross section negative [10, 40] in some range of transverse momentum kTk_T. The dipole amplitude obtained here does not display this issue.

Next we solve the leading order BK equation with fixed coupling, using the numerical data for N(r,x0)N(r,x_0) as an initial condition at x0=0.01x_0=0.01. Note that at this order in αs\alpha_s the coupling constant does not run in the LCPT calculation of the initial condition, and consequently we also limit ourselves to the fixed coupling case here. Evolution over 6 units of rapidity is shown in Fig. 4. For comparison, we also solve the BK equation using the modified MV-model initial condition with parameters as shown in Table I. This parameterized initial condition has a completely different behavior in the infrared region with N(r)→1N(r)\to1 at large rr whereas the numerical data gives a decreasing N(r)N(r) when rr exceeds a few fm, as already mentioned above. However, as can be seen in Fig. 4 the resulting BK-evolved dipole amplitudes are basically identical in the perturbative r≲0.5r\lesssim0.5 fm domain. In fact, due to the approach to the fixed point of the BK equation [41, 42, 43], at high rapidity the difference between the scattering amplitudes evolved with the two initial conditions diminishes. This demonstrates that the BK-evolved amplitude at small rr is not affected by the uncontrolled large-rr extrapolation of the initial condition.

One may define a saturation radius rsr_s, and a corresponding saturation momentum Qs=2/rsQ_s=\sqrt{2}/r_s from the condition that N(rs)=1−exp⁡(−12)≃0.4N(r_s)=1-\exp(-\frac{1}{2})\simeq0.4. For this to be a perturbative scale requires about 6 units of rapidity evolution, as can also be seen from Fig. 4. This corresponds to x≃2.5⋅10−5x\simeq2.5\cdot10^{-5}, where rs≃0.3r_s\simeq0.3 fm, and Qs≃1Q_s\simeq1 GeV. These values are not very far from the first “saturation model” fit to HERA DIS data by Golec-Biernat and Wüsthoff [44] from 25 years ago. Many more recent fits mentioned above have since confirmed that reaching the strong scattering regime with a small dipole and a proton target requires deep evolution to rather small xx. Also, some studies [17] of diffractive small-xx scattering of a qqˉ−gq\bar q-g state from the proton have indicated that the regime of ”color transparency” sets in when the typical transverse distance of the gluon from the qqˉq\bar q is around 0.2 fm.

Since scattering at x=0.01x=0.01 is fairly weak we have also evolved our initial condition with the linear BFKL equation [45, 46, 47], see Figure 5. After a few units of rapidity evolution, the linear equation begins to violate unitarity, N(r)≤1N(r) \leq1, at large rr. However, this regime of large dipoles is not under control in any case. More importantly though, at y=2−4y = 2-4 the absence of the nonlinear correction begins to affect the solution significantly even at rr substantially less than 1 fm. With BFKL we also noticed a greater difference between evolving the actual numerical data vs. the analytic modified MV-model parametrization (not shown), which differ in their large-rr extrapolation. Therefore, for accurate results it appears to be rather important to evolve with the non-linear BK equation even if one restricts to r<1r < 1 fm.

Figure 5. Leading-log BK (solid lines) vs. BFKL (dotted lines) evolution starting at x0=0.01x_0 = 0.01.

Let us finally study how the xx dependence obtained from the direct, fixed order NLO LCPT calculation differs to the one obtained by solving the BK equation. We note that in the LCPT calculation xx is an explicit cutoff for the longitudinal momentum of the emitted gluon, and this gluon emission is calculated in exact kinematics. On the other hand, in BK evolution multiple soft gluon emissions are resummed. This comparison is done by calculating the dipole amplitude at x=0.01x = 0.01 directly from the LCPT using Eq. (1), and comparing that to the dipole amplitude obtained by solving the BK equation with the initial condition computed at x0=0.05x_0 = 0.05. The resulting dipole amplitudes are shown in Figure 6. The most significant difference between the fixed order O(g2)O(g^2) LCPT amplitude and the BK evolved dipole is that the evolution equation decreases the anomalous dimension γ\gamma towards the asymptotic value γ∼0.6\gamma\sim0.6. On the other hand, the emission of one gluon in the direct LCPT calculation does not modify the extracted anomalous dimension, as can also be seen from Table I. This is, of course, the expected behavior. As already mentioned above, DIS phenomenology does not appear to support γ<1\gamma< 1 at x=0.01x = 0.01 or greater, so it seems reasonable to treat at least the emission of the first gluon with xg>0.01x_g > 0.01 explicitly in fixed order light-cone perturbation theory with exact kinematics.5

Figure 6. The dipole scattering amplitude at x=0.01x = 0.01 with and without prior BK evolution. The parameter γ\gamma denotes the resulting anomalous dimension fitted in the region 0.01 fm<r<0.5 fm0.01\,\mathrm{fm} < r < 0.5\,\mathrm{fm}.

Total cross section at small x

Next, we consider the DIS structure functions at small Bjorken-xx. The overall normalization of the dipole amplitude depends on the strong coupling constant αs=g2/(4π)\alpha_s = g^2/(4\pi), see Eq. (1). The same coupling constant also affects the Bjorken-xx dependence of the dipole amplitude via the BK evolution. In this work, our strategy is to fix the value of αs\alpha_s by calculating the total charm production cross section, and comparing it to the HERA reduced cross section data from Ref. [48]. We set the initial condition for the BK evolution at x0=0.01x_0 = 0.01, and compare to the HERA data in the region x<0.01x < 0.01, Q2<100 GeV2Q^2 < 100\,\mathrm{GeV}^2 (note that the smallest Q2Q^2 bin in the data is Q2=2.5 GeV2Q^2 = 2.5\,\mathrm{GeV}^2). In this region, there are N=39N = 39 data points. The experimental data is reported as reduced cross section

σr(x,y,Q2)=F2(x,Q2)−y21+(1−y)2FL(x,Q2).(6)\sigma_r(x,y,Q^2) = F_2(x,Q^2) - \frac{y^2}{1 + (1-y)^2} F_L(x,Q^2). \tag*{(6)}

Here y=Q2/(sx)y = Q^2/(sx) is the inelasticity variable, not to be confused with the evolution rapidity. The proton structure functions F2F_2 and FLF_L are expressed in terms of the total cross section for the virtual photon-proton cross section σγ∗p\sigma^{\gamma^* p}:

F2(x,Q2)=Q24παem(σTγ∗A+σLγ∗A)(7)F_2(x,Q^2) = \frac{Q^2}{4\pi\alpha_{\mathrm{em}}}\left(\sigma_T^{\gamma^* A}+\sigma_L^{\gamma^* A}\right) \tag*{(7)}
FL(x,Q2)=Q24παemσLγ∗A.(8)F_L(x,Q^2) = \frac{Q^2}{4\pi\alpha_{\mathrm{em}}}\sigma_L^{\gamma^* A}. \tag*{(8)}

In the dipole picture, the total cross section for the virtual photon-proton scattering can be expressed as a convolution of the photon wave function and the dipole amplitude as [2]

σT,Lγ∗A=2∑f∫d2b d2r dz∣Ψfγ∗→qqˉ(r,z,Q2)∣2N(r,b,xˉ).(9)\sigma_{T,L}^{\gamma^* A}=2\sum_f\int\mathrm{d}^2b\,\mathrm{d}^2r\,\mathrm{d}z\left|\Psi^{\gamma^*\to q\bar{q}}_f(r,z,Q^2)\right|^2N(r,b,\bar{x}). \tag*{(9)}

Here ff is the quark flavor, Q2Q^2 the photon virtuality and Ψfγ∗→qqˉ\Psi^{\gamma^*\to q\bar{q}}_f is the leading order wave function for the qqˉq\bar{q} Fock state of the virtual photon. In this equation we replace N(r,b,xˉ)N(r,b,\bar{x}) by the impact parameter averaged dipole amplitude N(r,xˉ)N(r,\bar{x}), as described above, and ∫d2b→ST\int\mathrm{d}^2b \to S_T. We fix the mass of the cc quark to 1.4 GeV1.4\ \mathrm{GeV}. The dipole amplitude in Eq. (9) is evaluated at xˉ=x(1+4mq2/Q2)\bar{x}=x(1+4m_q^2/Q^2), where mqm_q is the quark mass which enforces a smooth approach to the photoproduction limit [9, 44].

In order to confirm that the charm production cross section is not sensitive to non-perturbatively large dipoles we show in Figure 7 the fraction of the total cross section at x=0.0056x=0.0056, Q2=10 GeV2Q^2=10\ \mathrm{GeV}^2 as a function of the upper limit rmax⁡r_{\max} for the rr integral in Eq. (9). It is evident that the charm cross section is saturated by small dipoles whereas the inclusive cross section (calculated using mq=0.14 GeVm_q=0.14\ \mathrm{GeV} for the light quarks) at Q2=10 GeV2Q^2=10\ \mathrm{GeV}^2 is sensitive to larger dipoles beyond sizes where we may trust our calculation. When using a modified MV-model parameterization as an initial condition for the evolution with different extrapolation in the infrared region, one needs to integrate up to even larger rr in order to recover the full result for F2F_2. The charm production cross section is not affected by the different large-rr extrapolation (not shown). Qualitatively similar results have been obtained with the commonly used IPsat parameterization for the dipole-proton amplitude where, typically, even larger dipole sizes contribute as compared to the setup with factorized impact parameter dependence applied here [35, 49]. For these reasons we shall focus on charm production. In the future, our approach could be applied to other hard, perturbative processes such as, for example, single-inclusive production at high enough transverse momentum.

Figure 7. The fraction of the charm and inclusive DIS structure functions at x=0.0056x = 0.0056, Q2=10 GeV2Q^2 = 10\ \mathrm{GeV}^2 (corresponding to xˉ=0.01\bar{x} = 0.01 in the case of charm production) as function of the cutoff on the dipole size imposed in Eq. (9). The total cross section is the sum of light quark (mass 0.14 GeV0.14\ \mathrm{GeV}) and charm quark (mass 1.4 GeV1.4\ \mathrm{GeV}) production contributions.

Considering only the strong coupling constant αs\alpha_s as free parameter we obtain a reasonably good description of the charm production data. The value of χ2/(N−1)\chi^2/(N-1) as a function of αs\alpha_s is shown in Figure 8 using two different density profiles (Gaussian and hard sphere) for the proton. These two setups have different upper limits for the impact parameter bb and corresponding different transverse areas for the proton. The extracted optimal values for the strong coupling constant are αs=0.200\alpha_s=0.200 for the Gaussian proton and αs=0.181\alpha_s=0.181 for the hard sphere profile. These values are used throughout this paper. We note that fits of the total (rather than charm) cross section with tuned initial conditions [9, 10] and running coupling corrections to the BK equation have achieved lower χ2/N≈1\chi^2/N\approx1, without being able to simultaneously describe the charm data [9]. However, with our calculated initial condition there is room for the expected improvements of the photon wave function, evolution equation, and, of course, of the initial condition.

Figure 8. Determining the strong coupling constant by fitting the HERA charm production data in the region Q2<100 GeV2Q^2 < 100\ \mathrm{GeV}^2, x<0.01x < 0.01, with leading-log BK evolution started at x0=0.01x_0 = 0.01. The solid lines are polynomial fits to the computed values used to extract the optimal values for αs\alpha_s. The optimal values are αs=0.200\alpha_s = 0.200, χ2/(N−1)=2.27\chi^2/(N - 1) = 2.27 for the Gaussian proton and αs=0.181\alpha_s = 0.181, χ2/(N−1)=2.28\chi^2/(N - 1) = 2.28 for the hard sphere proton.

In this analysis, we have fixed the collinear regulator to mcoll=0.2m_{\mathrm{coll}} = 0.2 GeV in the LCPT calculation of the initial condition. As the charm cross section is dominated by small dipoles, our results are not highly sensitive to the actual value of this regulator: changing mcollm_{\mathrm{coll}} by a factor of 2 changes χ2/(N−1)\chi^2/(N-1) to HERA data by only 2 . . . 5% when using the optimal αs\alpha_s. We also keep the charm mass fixed to 1.4 GeV. The optimal value for αs\alpha_s and the achieved χ2/(N−1)\chi^2/(N-1) naturally will depend on this choice. We choose to work with fixed quark mass and collinear regulator and do not attempt to fit these simultaneously with αs\alpha_s, as the purpose of this work is to demonstrate the feasibility of computing the initial condition for the BK equation, and we emphasize that numerically potentially important higher order effects such as running coupling are still missing from the setup.

A comparison to the HERA charm production data in different Bjorken-xx bins is shown in Figure 9 as a function of the photon virtuality. We have checked that these results remain the same if we use the analytic parameterization (5), with parameters from table I, as initial condition; this confirms the insensitivity of the charm cross section to the large-rr extrapolation of the scattering amplitude.

Figure 9. Charm production reduced cross section at s=318 GeV\sqrt{s} = 318\,\mathrm{GeV} compared to the HERA data. Results are shown in the region where xˉ≤0.01\bar{x} \le0.01.

At x=0.008x = 0.008 there is only very little (≤0.2\le0.2 units of rapidity when x0=0.01x_0 = 0.01) evolution, so the dipole amplitude is almost completely determined by our initial condition. On the other hand, we also show results at lower xx which is dominated by the BK evolution. In addition to our standard setup where the initial condition for the BK evolution is set at x0=0.01x_0 = 0.01, we also show results using an initial condition computed at larger x0=0.05x_0 = 0.05. Note the weak dependence of this observable, at least, on where the “hand-off” from the xx-dependent initial condition to the evolution equation occurs. In contrast, ad-hoc initial condition parametrizations have to be re-tuned when x0x_0 is changed.

Figure 9 shows a fair agreement of the reduced cross section obtained from our light-cone wave function with the HERA charm data. Close to the initial condition we obtain a slightly slower Q2Q^2 dependence than seen in the data. As a result of the evolution this changes into faster virtuality dependence at very small xx. This is because the BK evolution at fixed coupling develops a small anomalous dimension γ≈0.6\gamma\approx0.6 for the dipole amplitude and a smaller anomalous dimension results in faster Q2Q^2 dependence.

Lastly, in Figure 10 we study how sensitive the charm production cross section is on the chosen proton density profile, and as such on the maximum impact parameter bmax⁡b_{\max} used in Eq. (3). The cross section is calculated at x=0.008x = 0.008 which is close to the initial condition for the BK evolution, again set at x0=0.01x_0 = 0.01. In both cases, we use the optimal value for the strong coupling constant extracted above. The cross section increases only slightly when the hard sphere profile with larger bmax⁡b_{\max} is used, but

Figure 10. Reduced cross section at s=318 GeV\sqrt{s} = 318\,\mathrm{GeV} calculated close to the initial condition using the Gaussian and Hard sphere density profiles, respectively.

the dependence on the virtuality is not affected. This weak dependence on the selected proton profile confirms that our results are not sensitive to non-perturbatively large impact parameters.

Summary and discussion

The present work represents a first attempt at relating the short-distance structure of the proton at high energies to its low-energy properties, covering several orders of magnitude in energy. We start from an effective three quark light-cone wave function which models the non-perturbative longitudinal and transverse momentum distributions of quarks at x≳0.1x \gtrsim0.1 as well as some of their correlations. The next step involves the computation, using exact kinematics, in light-cone perturbation theory of the O(g)O(g) correction to the light-cone wave function due to the emission of a gluon, and the O(g2)O(g^2) virtual corrections due to the exchange of a gluon by two quarks. This provides a leading twist contribution to the dipole scattering amplitude, N(r)∼r2γN(r) \sim r^2\gamma with γ=1\gamma= 1 at small rr. Optimistically, the LCPT correction extends the validity of the resulting light-cone wave function into the regime of perturbative transverse momenta, and parton momentum fractions x≳0.01x \gtrsim0.01. The corresponding dipole scattering amplitude is then evolved to yet higher energies (lower xx) by solving the BK equation, which resums emissions of additional soft gluons, and generates power corrections and an anomalous dimension.

The convolution of the LO photon light-cone wave function with the impact parameter averaged BK dipole scattering amplitude at leading logarithmic accuracy provides a fair description of the reduced DIS charm cross section measured at HERA, for αs≃0.2\alpha_s \simeq0.2. This value of the strong coupling was obtained from a fit to the charm cross section at Q2<100Q^2 < 100 GeV2^2 and x<0.01x < 0.01. None of the parameters of the low-energy three-quark model wave function were re-tuned to the high-energy data. Despite the fair description of the highly accurate data the resulting χ2/Ndof≈2.27\chi^2/N_{\mathrm{dof}} \approx2.27 with Ndof=38N_{\mathrm{dof}} = 38 implies a very low statistical significance, i.e. a very low probability that the data represents statistical fluctuations about the model predictions: the integral over the χ2\chi^2 distribution from χ2=2.27×38\chi^2 = 2.27 \times38 to infinity, commonly denoted as the ”p-value”, is 1.3×10−51.3 \times10^{-5}. However, the very low statistical significance of the fit should not be confused with a need for large corrections, Figure 9 shows that this is clearly not the case. This is entirely expected since there are multiple known sources of corrections such as, for example, of the photon wave function [50, 51, 52], of the evolution equation [53, 54, 55, 56, 57], and, of course, of the initial condition for the evolution equation (our proton light-cone wave function) which, e.g., may be improved with running coupling corrections. The data requires fairly moderate but systematic improvements of the model predictions across the relevant ranges of xx and Q2Q^2.

We have also provided analytic parameterizations of the impact parameter averaged dipole scattering amplitude for x=0.01…0.05x = 0.01 \ldots0.05 which accurately fit the numerical data in the regime of perturbative dipoles, r≲0.5r \lesssim0.5 fm. These parameterizations can be used in practice to estimate the corrections predicted by more accurate evolution equations. In the supplementary material we also provide the tabulated numerical data for N(r,x)N(r,x). Their large-rr extrapolation differs from that of the analytic parameterizations which allows for tests of the (in-)sensitivity to the uncontrolled non-perturbative regime of large dipoles. The quest for more accurate theoretical predictions at high energy for the upcoming EIC at BNL [58] and the proposed LHeC/FCC-he at CERN [59] requires initial conditions for the evolution equations which do not absorb theoretical improvements into a re-tune of their parameters.

Acknowledgments

This material is based upon work supported by the U.S. Department of Energy, Office of Science, Office of Nuclear Physics, within the framework of the Saturated Glue (SURGE) Topical Theory Collaboration. A.D. acknowledges support by the DOE Office of Nuclear Physics through Grant DE-SC0002307, and The City University of New York for PSC-CUNY Research grant 65079-00 53. This work was supported by the Academy of Finland, the Centre of Excellence in Quark Matter and projects 338263 and 346567 (H.M), and projects 347499 and 353772 (R.P). H.M is also supported under the European Union’s Horizon 2020 research and innovation programme by the European Research Council (ERC, grant agreement No. ERC-2018-ADG-835105 YoctoLHC) and by the STRONG-2020 project (grant agreement No. 824093). The content of this article does not reflect the official opinion of the European Union and responsibility for the information and views expressed therein lies entirely with the authors. Computing resources from CSC — IT Center for Science in Espoo, Finland and from the Finnish Grid and Cloud Infrastructure (persistent identifier urn:nbn:fi:research-infras-2016072533) were used in this work.

References

References

  1. [5]Y. V. Kovchegov, Small-x F₂ structure function of a nucleus including multiple pomeron exchanges, Phys. Rev. D 60 (1999) 034008 [arXiv:hep-ph/9901281].DOI
  2. [6]I. Balitsky, Operator expansion for high-energy scattering, Nucl. Phys. B 463 (1996) 99 [arXiv:hep-ph/9509348].DOI
  3. [7]F. Dominguez, C. Marquet, B.-W. Xiao and F. Yuan, Universality of Unintegrated Gluon Distributions at small x, Phys. Rev. D 83 (2011) 105005 [arXiv:1101.0715 [hep-ph]].DOI
  4. [8]J. L. Albacete, N. Armesto, J. G. Milhano and C. A. Salgado, Non-linear QCD meets data: A Global analysis of lepton-proton scattering with running coupling BK evolution, Phys. Rev. D 80 (2009) 034031 [arXiv:0902.1112 [hep-ph]].
  5. [9]J. L. Albacete, N. Armesto, J. G. Milhano, P. Quiroga-Arias and C. A. Salgado, AAMQS: A non-linear QCD analysis of new HERA data at small-x including heavy quarks, Eur. Phys. J. C 71 (2011) 1705 [arXiv:1012.4408 [hep-ph]].
  6. [10]T. Lappi and H. Mäntysaari, Single inclusive particle production at high energy from HERA data to proton-nucleus collisions, Phys. Rev. D 88 (2013) 114020 [arXiv:1309.6963 [hep-ph]].
  7. [11]I. Balitsky, Quark contribution to the small-x evolution of color dipole, Phys. Rev. D 75 (2007) 014001 [arXiv:hep-ph/0609105].DOI
  8. [12]G. Beuf, H. Hänninen, T. Lappi and H. Mäntysaari, Color Glass Condensate at next-to-leading order meets HERA data, Phys. Rev. D 102 (2020) 074028 [arXiv:2007.01645 [hep-ph]].DOI
  9. [13]H. Hänninen, H. Mäntysaari, R. Paatelainen and J. Penttala, Proton Structure Functions at Next-to-Leading Order in the Dipole Picture with Massive Quarks, Phys. Rev. Lett. 130 (2023) no. 19 192301 [arXiv:2211.03504 [hep-ph]].
  10. [14]A. Dumitru and R. Paatelainen, Sub-femtometer scale color charge fluctuations in a proton made of three quarks and a gluon, Phys. Rev. D 103 (2021) no. 3 034026 [arXiv:2010.11245 [hep-ph]].DOI
  11. [15]A. Dumitru, H. Mäntysaari and R. Paatelainen, Color charge correlations in the proton at NLO: Beyond geometry based intuition, Phys. Lett. B 820 (2021) 136560 [arXiv:2103.11682 [hep-ph]].DOI
  12. [16]F. Hautmann, Z. Kunszt and D. E. Soper, Hard scattering factorization and light cone Hamiltonian approach to diffractive processes, Nucl. Phys. B 563 (1999) 153 [arXiv:hep-ph/9906284].DOI
  13. [17]F. Hautmann and D. E. Soper, Color transparency in deeply inelastic diffraction, Phys. Rev. D 63 (2001) 011501 [arXiv:hep-ph/0008224].
  14. [18]F. Hautmann, Power-like corrections and the determination of the gluon distribution, Phys. Lett. B 643 (2006) 171 [arXiv:hep-ph/0610078].
  15. [19]F. Hautmann and D. E. Soper, Parton distribution function for quarks in an s-channel approach, Phys. Rev. D 75 (2007) 074020 [arXiv:hep-ph/0702077].DOI
  16. [20]X. Ji, Y.-S. Liu, Y. Liu, J.-H. Zhang and Y. Zhao, Large-momentum effective theory, Rev. Mod. Phys. 93 (2021) no. 3 035005 [arXiv:2004.03543 [hep-ph]].DOI
  17. [21]X. Ji and Y. Liu, Computing light-front wave functions without light-front quantization: A large-momentum effective theory approach, Phys. Rev. D 105 (2022) no. 7 076014 [arXiv:2106.05310 [hep-ph]].
  18. [23]B. Pasquini, S. Rodini and S. Venturini, The valence quark, sea, and gluon content of the pion from the parton distribution functions and the electromagnetic form factor, arXiv:2303.01789 [hep-ph].DOI
  19. [24]F. Schlumpf, Relativistic constituent quark model of electroweak properties of baryons, Phys. Rev. D 47 (1993) 4114 [arXiv:hep-ph/9212250]. [Erratum: Phys.Rev.D 49, 6246 (1994).]arxiv.org/abs/hep-ph/9212250
  20. [25]S. J. Brodsky and F. Schlumpf, Wave function independent relations between the nucleon axial coupling g_A and the nucleon magnetic moments, Phys. Lett. B 329 (1994) 111 [arXiv:hep-ph/9402214].
  21. [27]A. Dumitru, H. Mäntysaari and R. Paatelainen, Cubic color charge correlator in a proton made of three quarks and a gluon, Phys. Rev. D 105 (2022) no. 3 036007 [arXiv:2106.12623 [hep-ph]].DOI
  22. [28]A. Dumitru, H. Mäntysaari and R. Paatelainen, Stronger C-odd color charge correlations in the proton at higher energy, Phys. Rev. D 107 (2023) no. 1 L011501 [arXiv:2210.05390 [hep-ph]].DOI
  23. [29]A. Dumitru and T. Stebel, Multiquark matrix elements in the proton and three gluon exchange for exclusive ρ_c production in photon-proton diffractive scattering, Phys. Rev. D 99 (2019) no. 9 094038 [arXiv:1903.07660 [hep-ph]].DOI
  24. [30]J. Berger and A. M. Stasto, Small x nonlinear evolution with impact parameter and the structure function data, Phys. Rev. D 84 (2011) 094022 [arXiv:1106.5740 [hep-ph]].DOI
  25. [31]H. Mäntysaari and B. Schenke, Confronting impact parameter dependent JIMWLK evolution with HERA data, Phys. Rev. D 98 (2018) no. 3 034013 [arXiv:1806.06783 [hep-ph]].DOI
  26. [32]H1 collaboration, A. Aktas et al., Elastic J/ψ production at HERA, Eur. Phys. J. C 46 (2006) 585 [arXiv:hep-ex/0510016].DOI
  27. [33]H. Kowalski, L. Motyka and G. Watt, Exclusive diffractive processes at HERA within the dipole picture, Phys. Rev. D 74 (2006) 074016 [arXiv:hep-ph/0606272].DOI
  28. [34]A. Dumitru, H. Mäntysaari and R. Paatelainen, Impact parameter dependence of color charge correlations in the proton, SciPost Phys. Proc. 8 (2022) 058 [arXiv:2105.08503 [hep-ph]].arxiv.org/abs/2105.08503
  29. [35]H. Kowalski and D. Teaney, An Impact parameter dipole saturation model, Phys. Rev. D 68 (2003) 114005 [arXiv:hep-ph/0304189].DOI
  30. [36]L. D. McLerran and R. Venugopalan, Computing quark and gluon distribution functions for very large nuclei, Phys. Rev. D 49 (1994) 2233 [arXiv:hep-ph/9309289].DOI
  31. [37]H1 and ZEUS collaborations, F. D. Aaron et. al., Combined Measurement and QCD Analysis of the Inclusive e⁺e⁻ Scattering Cross Sections at HERA, JHEP 01 (2010) 109 [arXiv:0911.0884 [hep-ex]].
  32. [38]Y. V. Kovchegov and A. H. Mueller, Gluon production in current nucleus and nucleon - nucleus collisions in a quasiclassical approximation, Nucl. Phys. B 529 (1998) 451 [arXiv:hep-ph/9802440].
  33. [39]D. Kharzeev, Y. V. Kovchegov and K. Tuchin, Cronin effect and high pT suppression in pA collisions, Phys. Rev. D 68 (2003) 094013 [arXiv:hep-ph/0307037].DOI
  34. [40]B. Ducloué, T. Lappi and Y. Zhu, Implementation of NLO high energy factorization in single inclusive forward hadron production, Phys. Rev. D 95 (2017) no. 11 114007 [arXiv:1703.04962 [hep-ph]].DOI
  35. [41]A. M. Stasto, K. J. Golec-Biernat and J. Kwiecinski, Geometric scaling for the total γ* p cross-section in the low x region, Phys. Rev. Lett. 86 (2001) 596 [arXiv:hep-ph/0007192].DOI
  36. [42]S. Munier and R. B. Peschanski, Geometric scaling as traveling waves, Phys. Rev. Lett. 91 (2003) 232001 [arXiv:hep-ph/0309177].DOI
  37. [43]S. Munier and R. B. Peschanski, Traveling wave fronts and the transition to saturation, Phys. Rev. D 69 (2004) 034008 [arXiv:hep-ph/0310357].DOI
  38. [44]K. J. Golec-Biernat and M. Wusthoff, Saturation effects in deep inelastic scattering at low Q² and its implications on diffraction, Phys. Rev. D 59 (1998) 014017 [arXiv:hep-ph/9807513].
  39. [45]L. N. Lipatov, Reggeization of the Vector Meson and the Vacuum Singularity in Nonabelian Gauge Theories, Sov. J. Nucl. Phys. 23 (1976) 338.
  40. [46]E. A. Kuraev, L. N. Lipatov and V. S. Fadin, The Pomeranchuk Singularity in Nonabelian Gauge Theories, Sov. Phys. JETP 45 (1977) 199.
  41. [47]I. I. Balitsky and L. N. Lipatov, The Pomeranchuk Singularity in Quantum Chromodynamics, Sov. J. Nucl. Phys. 28 (1978) 822.
  42. [48]H1 and ZEUS collaborations, H. Abramowicz et. al., Combination and QCD analysis of charm and beauty production cross-section measurements in deep inelastic ep scattering at HERA, Eur. Phys. J. C 78 (2018) no. 6 473 [arXiv:1804.01019 [hep-ex]].arxiv.org/abs/1804.01019
  43. [49]H. Mäntysaari and P. Zurita, In depth analysis of the combined HERA data in the dipole models with and without saturation, Phys. Rev. D 98 (2018) 036002 [arXiv:1804.05311 [hep-ph]].DOI
  44. [50]G. Beuf, T. Lappi and R. Paatelainen, Massive Quarks at One Loop in the Dipole Picture of Deep Inelastic Scattering, Phys. Rev. Lett. 129 (2022) no. 7 072001 [arXiv:2112.03158 [hep-ph]].DOI
  45. [51]G. Beuf, T. Lappi and R. Paatelainen, Massive quarks in NLO dipole factorization for DIS: Longitudinal photon, Phys. Rev. D 104 (2021) no. 5 056032 [arXiv:2103.15459 [hep-ph]].
  46. [52]G. Beuf, T. Lappi and R. Paatelainen, Massive quarks in NLO dipole factorization for DIS: Transverse photon, Phys. Rev. D 106 (2022) no. 3 034013 [arXiv:2204.02486 [hep-ph]].DOI
  47. [53]T. Lappi and H. Mäntysaari, Next-to-leading order Balitsky-Kovchegov equation with resummation, Phys. Rev. D 93 (2016) no. 9 094004 [arXiv:1601.06598 [hep-ph]].DOI
  48. [54]I. Balitsky and G. A. Chirilli, Next-to-leading order evolution of color dipoles, Phys. Rev. D 77 (2008) 014019 [arXiv:0710.4330 [hep-ph]].
  49. [55]E. Iancu, J. D. Madrigal, A. H. Mueller, G. Soyez and D. N. Triantafyllopoulos, Resumming double logarithms in the QCD evolution of color dipoles, Phys. Lett. B 744 (2015) 293 [arXiv:1502.05642 [hep-ph]].DOI
  50. [56]B. Ducloué, E. Iancu, A. H. Mueller, G. Soyez and D. N. Triantafyllopoulos, Non-linear evolution in QCD at high-energy beyond leading order, JHEP 04 (2019) 081 [arXiv:1902.06637 [hep-ph]].DOI
  51. [57]E. Iancu, J. D. Madrigal, A. H. Mueller, G. Soyez and D. N. Triantafyllopoulos, Collinearly-improved BK evolution meets the HERA data, Phys. Lett. B 750 (2015) 643 [arXiv:1507.03651 [hep-ph]].DOI
  52. [58]R. Abdul Khalek et. al., Science Requirements and Detector Concepts for the Electron-Ion Collider: EIC Yellow Report, Nucl. Phys. A 1026 (2022) 122447 [arXiv:2103.05419 [physics.ins-det]].DOI
  53. [59]LHeC, FCC-he Study Group collaboration, P. Agostini et. al., The Large Hadron-Electron Collider at the HL-LHC, J. Phys. G 48 (2021) no. 11 110501 [arXiv:2007.14491 [hep-ex]].

Paper details

vlnc:W_24mhxqk8
Version
v1 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC BY 4.0
Usage

views · 0

downloads · 0

A view counts once per browser per day, a download each time the PDF is requested.

  • Physics and Astronomy

Not recorded for this paper

View statement
Looking for related papers…
Contents