betaFeedback

WorksPhysics and Astronomy

Quarkonium spin alignment in a vortical medium

Abstract

We use a potential model to investigate the phenomenology of quarkonium in a thermal rotating medium, where vorticity and spin density are not necessarily in equilibrium. We find that the quarkonium spin density matrix, as well as the binding energy and melting temperature, are sensitive to both the vorticity and the lack of equilibrium between vorticity and spin. This means that quarkonium spin alignment is a sensitive probe for vorticity and spin within the hydrodynamic phase. Information unequivocably pointing to spin-orbit non-equilibrium dynamics can be obtained from a combined study of quarkonium relative abundance and spin alignment, as well as experimentally obtainable off-diagonal density matrix elements.

Introduction

Quarkonium as been used as a probe of thermodynamic properties of the Quark-Gluon Plasma (QGP) since the seminal paper from [1]. The heavy mass scale means that it is a probe that can both be examined reliably by theoretical calculations and susceptible to non-trivial in-medium effects, serving as a “thermometer”, sensitive to the interplay of thermal fluctuations and the QCD medium, and as a probe for non-equilibrium behavior [4, 5] [2, 3].

The advent of the study of vorticity in heavy ion collisions [6] added a potentially new arena where quarkonium could be used. In fact, we shall argue that quarkonium provides unique opportunities for the phenomenology of hydrodynamics with spin. First of all, quarkonium can be formed early in the collision and can survive throughout the quark gluon plasma evolution. Unlike polarized Λs [6] and spin-aligned vector mesons [7, 8, 9, 10, 11, 12], they are potentially sensitive to the entire dynamics of the fluid and not just to the freeze-out.

Furthermore, the long-lived quarkonium state is spin 1, having a 3×33 \times3 density matrix with 8 degrees of freedom, 6 of which are accessible to spin alignment measurements. Thus, if, as seems to be theoretically highly likely [13, 14, 15, 16, 17, 18, 19, 20], spin and vorticity are not in equilibrium, this lack of equilibration will can be imprinted on the density matrix’s measurable off-diagonal elements ρ0,±1,ρ±1,∓1\rho_{0,\pm1}, \rho_{\pm1,\mp1} [11] (No equivalent elements exist for the “qubit” density matrix of a fermion). In fact, these elements have recently been measured [21].

Last but not least, bottomonium and to a certain extent charmonium states can be viewed as solutions to a Schrodinger equation with heavy Quark wavefunctions moving around a QCD potential (including a weakly coupled and a confined part) [22, 23]. It is therefore possible to understand, both analytically [24] and via effective theory [25], effect that vorticity will have to the properties of quarkonium.

In this work, we shall go in this direction, combining the insights developed in [11] with a potential model for quarkonium solved using standard methods extended to rotating frames [26, 27, 28, 29] and finite temperature [30, 31]. In section II we shall assess currently available experimental data, and point out what would be necessary to probe spin-orbit non-equilibrium. Then in § 1 we calculate the quarkonium wavefunction properties, namely binding energy and a semi-classical estimate of the melting temperature, in a rotating frame. Finally, in section IV we calculate quarkonium observables that could indicate a lack of equilibrium between spin and vorticity.

The Quarkonium spin density matrix

In [11], we have argued that vector spin alignment contains crucial information on the still-unknown spin hydrodynamic evolution in heavy ion collisions, provided that not just the ρ00\rho_{00} coefficient but the “off-diagonal” coefficients are measured. The former, ρ00\rho_{00} was measured in [7, 8] and is sensitive to θ\theta, the angle w.r.t. the spin alignment direction, which for heavy ions would be mainly the reaction plane, reflecting the vorticity structure in analogy to the global polarization measurement [6].

The other coefficients, related to density matrix elements ρ0±1,±1±1,±1∓1\rho_{0\pm1,\pm1\pm1,\pm1\mp1} (see equation (1) of [11]) would depend on a second “reference” angle ϕ\phi, whose most logical definition is in terms of the beam axis.

In such a set-up, the non-equilibrium between spin and vorticity would manifest itself by the interplay of transverse vs. longitudinal polarization [6]. The heavy ion system actually lends hope that this interplay could occur (Figure 1): Transverse polarization is thought to be present from the beginning, since it is the result of the angular momentum of the initial state of an off-central collision. Longitudinal polarization forms from azimuthal gradients, on a time-scale comparable to the formation of momentum anisotropy causing elliptic flow. If polarization “relaxes to vorticity” [13, 15], one expects polarization to be aligned in the transverse direction, but comparatively mis-aligned in the longitudinal one [15], which indeed seems to be out of phase with some dynamical calculations [6]. Hence, there will be an angle in the z−yz-y plane between vorticity and polarization, whose magnitude we do not yet know to estimate quantitatively but whose experimental confirmation will add a phenomenological dimension to the recent theoretical development of spin hydrodynamics [13, 14, 15, 16, 17, 18, 19, 20].

Figure 1. A schematic representation of how the vorticity in heavy ion collisions looks like at early and late times. Blue dashed arrows represent vorticity, cyan dot-dashed ones polarization, red solid ones flow. The mis-alignment of spin in the right panel is due to the lack of equilibrium between vorticity and polarization [13, 15].

The hope, therefore, is that off-diagonal (longitudinal-transverse) coefficients are potentially important for they could signal deviations from local equilibrium (included via a Cooper-Frye type formula [32]) due to the presence of two distinct axial currents, representing spin and vorticity, evolving on different time-scales [15]. In [11] we illustrated this with a coalescence type model.

Coalescence of only spin within a vortical background should not change the coherence of the density matrix, since it is a unitary process and the dynamics is symmetric around the vortical axis. But assuming vorticity and pre-existing spin density are not in equilibrium and pointing in different directions, this is no longer true [11]; vorticity is “classical” background, interacting with the quantum spin state, so if coalescence happens in a vortical background (i.e. if spin and vorticity are out of equilibrium) one expects impurity of the density matrix. Mathematically, the loss of purity is manifest in Eq. (23) of [11] PL(ω)P_L(\omega) representing the (unknown) classical probability of a vortex ω\omega giving an angular momentum LL to the meson wavefunction. When this probability becomes uniform (PL(ω)→constantP_L(\omega) \to\text{constant}) we recover a maximally impure state1. For vector bosons, this impurity is manifest in the off-diagonal

matrix components (see the discussion between eqs 4 and 5 of [11]).

While this data as yet does not exist for vector mesons, it does exist for quarkonia J/ψJ/\psi and Υ\Upsilon states [21], since the ALICE Collaboration measurement of the quarkonium polarization included the off-diagonal values of the spin density matrix. Therefore, we can do direct connection between polarization parameters λθ\lambda_\theta, λϕ\lambda_\phi and λθϕ\lambda_{\theta\phi} and density matrix [21] and the coefficients used in [11] (table in eq. 1)

VariableElementcoefficient ×34π\times\frac{3}{4\pi}
ρ00\rho_{00}ρ00\rho_{00}cos⁡2θ\cos^{2}\theta
1−ρ002\frac{1-\rho_{00}}{2}ρ11+ρ−1−12\frac{\rho_{11}+\rho_{-1-1}}{2}sin⁡2θ\sin^{2}\theta
r10r_{10}Re⁡[ρ−10−ρ10]\operatorname{Re}[\rho_{-10}-\rho_{10}]sin⁡(2θ)cos⁡(ϕ)\sin(2\theta)\cos(\phi)
α10\alpha_{10}Im⁡[−ρ−10+ρ10]\operatorname{Im}[-\rho_{-10}+\rho_{10}]sin⁡(2θ)sin⁡(ϕ)\sin(2\theta)\sin(\phi)
r1,−1r_{1,-1}Re⁡[ρ1,−1]\operatorname{Re}[\rho_{1,-1}]sin⁡2θcos⁡(2ϕ)\sin^{2}\theta\cos(2\phi)
α1,−1\alpha_{1,-1}Im⁡[ρ1,−1]\operatorname{Im}[\rho_{1,-1}]sin⁡2θsin⁡(2ϕ)\sin^{2}\theta\sin(2\phi)

Table 2.

ρ00=1+λθ3+λθ,r1,−1=λϕ3+λθ,r10=λθϕ3+λθ(1)\rho_{00} = \frac{1+\lambda_\theta}{3+\lambda_\theta}, \qquad r_{1,-1} = \frac{\lambda_\phi}{3+\lambda_\theta}, \qquad r_{10} = \frac{\lambda_{\theta\phi}}{3+\lambda_\theta} \tag*{(1)}

This is possible by making the comparison between J/ψJ/\psi angular distribution and the standard vector mesons angular distribution shown in Eq. 1 of [11] (Eq. (1)).

Thus, we can do the analyses presented in [11] to relate λθ,ϕ,θϕ\lambda_{\theta,\phi,\theta\phi} to the wave function coherence via the parametrization in terms of Gell-Mann matrices. Choosing the n3−8n_{3-8} basis for this parametrization, we need to solve the following system of algebraic equations derived in [11] in terms of the frame relating the lab to the spin direction (defined by angles θr,ϕr\theta_r,\phi_r)

112(3(n8−3n3)cos⁡(2θr)−3n3+n8+4)=ρ00(2)\frac{1}{12}\left(3\left(n_8-\sqrt{3}n_3\right)\cos(2\theta_r)-\sqrt{3}n_3+n_8+4\right)=\rho_{00} \tag*{(2)}
(n8−3n3)sin⁡(θr)cos⁡(θr)cos⁡(ϕr)2=r10(3)\frac{\left(n_8-\sqrt{3}n_3\right)\sin(\theta_r)\cos(\theta_r)\cos(\phi_r)}{\sqrt{2}}=r_{10} \tag*{(3)}
−(3n3+3n8)sin⁡(θr)sin⁡(ϕr)32=α10(4)-\frac{\left(\sqrt{3}n_3+3n_8\right)\sin(\theta_r)\sin(\phi_r)}{3\sqrt{2}}=\alpha_{10} \tag*{(4)}
ϕr=−12tan⁡−1(α1,−1r1,−1)(5)\phi_r=-\frac{1}{2}\tan^{-1}\left(\frac{\alpha_{1,-1}}{r_{1,-1}}\right) \tag*{(5)}

Now, we will do the follow change variable n~=n8−3n3\tilde{n}=n_8-\sqrt{3}n_3, and knowing that variables are equal to zero α10\alpha_{10} and α1,−1\alpha_{1,-1}. So, we can write this system equation the following form:

112(3n~cos⁡(2θr)+n~+4)=ρ00(6)\frac{1}{12}\left(3\tilde{n}\cos(2\theta_r)+\tilde{n}+4\right)=\rho_{00} \tag*{(6)}
n~sin⁡(θr)cos⁡(θr)cos⁡(ϕr)2=r10(7)\frac{\tilde{n}\sin(\theta_r)\cos(\theta_r)\cos(\phi_r)}{\sqrt{2}}=r_{10} \tag*{(7)}
ϕr=0(8)\phi_r=0 \tag*{(8)}
Equation not extracted
(9)

Therefore, we have the following solution:

n~(ρ00,r10)=−(1−3ρ00)2+3(1−3ρ00)4+4(1−3ρ00)2r1026ρ00−2(10)\tilde{n}(\rho_{00},r_{10})=-\frac{(1-3\rho_{00})^2+3\sqrt{(1-3\rho_{00})^4+4(1-3\rho_{00})^2r_{10}^2}}{6\rho_{00}-2} \tag*{(10)}
Θ1(ρ00,r10)=−2(1−3ρ00)2−2(1−3ρ00)4+4(1−3ρ00)2r102+6r1022(1−3ρ00)2+9r102(11)\Theta_1(\rho_{00},r_{10})=-\sqrt{\frac{2(1-3\rho_{00})^2-2\sqrt{(1-3\rho_{00})^4+4(1-3\rho_{00})^2r_{10}^2+6r_{10}^2}}{2(1-3\rho_{00})^2+9r_{10}^2}} \tag*{(11)}
Θ2(ρ00,r10)=(1−3ρ00)2−(1−3ρ00)4+4(1−3ρ00)2r102+3r1022(1−3ρ00)2+9r102(12)\Theta_2(\rho_{00},r_{10})=\sqrt{\frac{(1-3\rho_{00})^2-\sqrt{(1-3\rho_{00})^4+4(1-3\rho_{00})^2r_{10}^2+3r_{10}^2}}{2(1-3\rho_{00})^2+9r_{10}^2}} \tag*{(12)}
Θ3(ρ00,r10)=Θ2(ρ00,r10)(2(1−3ρ00)2+2(1−3ρ00)4+4(1−3ρ00)2r102)2(3ρ00−1)r10(13)\Theta_3(\rho_{00},r_{10})=\frac{\Theta_2(\rho_{00},r_{10})\left(2(1-3\rho_{00})^2+2\sqrt{(1-3\rho_{00})^4+4(1-3\rho_{00})^2r_{10}^2}\right)}{2(3\rho_{00}-1)r_{10}} \tag*{(13)}
θr=tan⁡−1(Θ1(ρ00,r10),Θ3(ρ00,r10))(14)\theta_r=\tan^{-1}\left(\Theta_1(\rho_{00},r_{10}),\Theta_3(\rho_{00},r_{10})\right) \tag*{(14)}

So, using polarization parameters that were obtained from ALICE collaboration [21] at different transverse momentum (pTp_T) ranges, we can determine whether the density matrix represents a coherence state or not. To make it we will use the equation 11 and reach the following figure 2.

Now, the coefficients in the frame Collins-Soper frame, given in the ranges 4<pT<64<p_T<6 (GeV/C2GeV/C^2) and 6<pT<106<p_T<10 (GeV/C2GeV/C^2) respectively result in n~=(0.09±0.10,0.02±0.07)\tilde{n}=(0.09\pm0.10,0.02\pm0.07). Making the comparison with the Helicity frame value n~=0.09±0.11\tilde{n}=0.09\pm0.11 to 6<pT<106<p_T<10 (GeV/C2GeV/C^2). Therefore, we can see that they are the same within error bar.

Looking at the figure 2, we can conclude that the density matrix from J/ψJ/\psi particle does not represent the pure state since none of the values for n3,8n_{3,8} obtained from the data intersects the black pink points, i.e. these points represent the pure state in other words when the density matrix satisfy ρ2=ρ\rho^2=\rho. This might indicate that statistical freeze-out advocated in

[32] [3] is a good estimate of particle production in heavy-ion collisions because the density matrix does not represent a coherent state as argued in [11]. In the bottomonium Υ(1S)\Upsilon(1S) case, we can see in the right panel of Figure 2 that because of large uncertainty do not know whether for these particles the density matrix represents a coherence state or not.

Figure 2. The experimental results of J/ψJ/\psi and Υ\Upsilon polarization measurements analyzed in terms of the Gell-Mann matrix representation of [11]. These results were made using the helicity and Collins-Soper frame to J/ψJ/\psi polarization [21]. The other Collins-Soper values was not take in this plot because they were above the Helicity 6<pT<106 < p_T < 10 (GeV/C2GeV/C^2) uncertainty band as were shown in the text.

To fully assess the significance of the estimate above one must recall Fig. 4 of [11] and the definition of PL(w)P_L(w) given in the introduction. We have no idea what PL(w)P_L(w) is beyond the fact that it overall conserves momentum, but it acts as a projector. One recovers a pure state when ρij=δLl\rho_{ij} = \delta_{Ll}, (there is a certainty of vorticity giving a certain momentum) and a maximally mixed state when the momentum given by vorticity is independent of LL. So the measurement in Fig. 4 [11] is directly connected to how out of equilibrium vorticity and spin are, and how much vorticity vs pre-existing spin influences the final spin of the vector meson. Linear combination of the different LL-values in Figure 4 of [11] are possible, illustrating a probability of different spin configurations.

Figure 4. Charmonium mass in relation to parameter CC for various energy and momentum levels (notation follows atomic physics quantum numbers) using T=168T = 168 MeV

Note that these coefficients are given in terms of an angle θ\theta, which in [11] is related to θr\theta_r, the angle between the hadronization frame and the lab frame. This angle of course depends on the detailed hydrodynamical and spin-hydrodynamical evolution of the system, but it is obviously highly dependent on the reaction plane angle Φ\Phi. Note that since local polarization necessarily averages to zero in symmetric collisions by parity, so must the off-diagonal matrix elements (unless large event by event fluctuations in transparency occur, something never seen in experimental data). Because of this, a zero finding integrated over azimuthal anisotropy can not count as conclusive evidence.

Considering the Harmonic behavior of the coefficients in Figure 4 of [11] w.r.t. θr\theta_r (most coefficients average to zero for all angles), therefore, it would be crucial to measure λθ,ϕ,θϕ\lambda_{\theta,\phi,\theta\phi} not as a function of pTp_T as in [21] but as function of azimuthal reaction plane angle (This can be measured directly at lower energies [36], and via cumulants at the LHC [37]). A modulated behavior would be a clear signature for a non-trivial PL(w)P_L(w) which can then be harmonically decomposed into L=0,1,2L=0,1,2 components of Figure 4 of [11] to obtain information of the impact of spin vs vorticity in J/ψJ/\psi vs Υ\Upsilon hadronization. If the dependence w.r.t. θr\theta_r will be compatible with zero as it was for pTp_T in each Φ\Phi bin, this is good evidence for a statistical Cooper-Frye freeze-out as in [32] [3]. Schematically, these two alternatives are illustrated in Figure 3.

Figure 3. A schematic illustration of how we expect the coefficients λθ,ϕ,θϕ\lambda_{\theta,\phi,\theta\phi} to evolve with reaction plane angle Φ−Φreaction\Phi-\Phi_{\mathrm{reaction}} in the two scenarios. In [21], these coefficients are compatible with zero when integrated over this angle. A flat dependence (a) indicates a maximally incoherent Cooper-Frye/thermal type production. A harmonic dependence (b) might indicate a more non-trivial coalescence scenario involving vorticity. Note that both cases are consistent with zero when integrated over Φ\Phi, as was done in [21]. NB: The exact azimuthal shape will need more realistic modeling, but the amplitude can be qualitatively estimated in Table I using [38, 39]

The quantitative details of Figure 3 would require a hydrodynamic simulation with a hydrodynamic model where spin and vorticity are not in equilibrium, given the interplay between spin, vorticity, radial flow and its anisotropies in the quarkonia pTp_T and rapidity distributions. We have used an analytically solvable model subject to ongoing work [38, 39] to provide an order of magnitude estimate of the maximum of the effect compatible with Figure 2. For reasons expanded on in the appendix this should be regarded as just that, an order of magnitude estimate, although it is gratifying that the possible off-diagonal terms are comparable in magnitude to λθ\lambda_\theta.

For this reason, in the rest of the paper we proceed to examine the microscopic dynamics of the charmonium state using a potential model, with a view of developing quantitative signatures of non-equilibrium between charmonium spin and vorticity.

The quarkonium state in rotating reference frames

The Schrödinger equation

We are interested in the angular momentum due to vortices couples with the quark spin. So, we need to take an extra term, ω⋅S\boldsymbol{\omega}\cdot\mathbf{S} [28],

H=(p−mω×r)22m−m2(ω×r)2−ω⋅S+V(r)(15)\mathcal{H}=\frac{(\mathbf{p}-m\boldsymbol{\omega}\times\mathbf{r})^2}{2m}-\frac{m}{2}(\boldsymbol{\omega}\times\mathbf{r})^2-\boldsymbol{\omega}\cdot\mathbf{S}+V(\mathbf{r}) \tag*{(15)}
λθ\lambda_\thetaλϕ\lambda_\phiλθϕ\lambda_{\theta\phi}
J/ψJ/\psi0.0210.0230.012
Υ\Upsilon0.0540.0540.002

Table 1. These are estimates of density matrix coefficients using Blast Wave Model [38, 39] more details in Appendix Appendix B .

Now, we can write the equation 16 for a two-body case. Thus we have the following expression:

H=∑i=1,2((pi−miωi×ri)22mi−mi2(ωi×ri)2−ωi⋅Si)+V(∣r1−r2∣)(16)\mathcal{H} = \sum_{i=1,2} \left( \frac{(\mathbf{p}_i - m_i \boldsymbol{\omega}_i \times\mathbf{r}_i)^2}{2m_i} - \frac{m_i}{2}(\boldsymbol{\omega}_i \times\mathbf{r}_i)^2 - \boldsymbol{\omega}_i \cdot\mathbf{S}_i \right) + V(|\mathbf{r}_1-\mathbf{r}_2|) \tag*{(16)}

Using the relations:

{P=p1+p2,p=μ(p1m1−p2m2),μ=m1m2m1+m2,r1=R+m2m1+m2r,r2=R−m1m1+m2r(17)\left\{ \begin{aligned} \mathbf{P} &= \mathbf{p}_1 + \mathbf{p}_2, \qquad\mathbf{p} = \mu\left(\frac{\mathbf{p}_1}{m_1} - \frac{\mathbf{p}_2}{m_2}\right),\\ \mu&= \frac{m_1m_2}{m_1+m_2},\\ \mathbf{r}_1 &= \mathbf{R} + \frac{m_2}{m_1+m_2}\mathbf{r}, \qquad\mathbf{r}_2 = \mathbf{R} - \frac{m_1}{m_1+m_2}\mathbf{r} \end{aligned} \right. \tag*{(17)}

As the two quarks that form the mesons are in the same vortical background, we can suppose that ω1=ω2=ω\boldsymbol{\omega}_1 = \boldsymbol{\omega}_2 = \boldsymbol{\omega} and S=S1+S2\mathbf{S} = \mathbf{S}_1 + \mathbf{S}_2, and then write:

H=P22M+p22μ−P⋅(ω×R)−p⋅(ω×r)−ω⋅(S1+S2)+V(∣r∣)(18)\mathcal{H} = \frac{P^2}{2M} + \frac{p^2}{2\mu} - \mathbf{P} \cdot(\boldsymbol{\omega} \times\mathbf{R}) - \mathbf{p} \cdot(\boldsymbol{\omega} \times\mathbf{r}) - \boldsymbol{\omega} \cdot(\mathbf{S}_1+\mathbf{S}_2) + V(|\mathbf{r}|) \tag*{(18)}

Since S=S1+S2\mathbf{S} = \mathbf{S}_1 + \mathbf{S}_2 and we are interested just in the reduced coordinate,

H=p22μ−p⋅(ω×r)−ω⋅S+V(∣r∣)(19)\mathcal{H} = \frac{p^2}{2\mu} - \mathbf{p} \cdot(\boldsymbol{\omega} \times\mathbf{r}) - \boldsymbol{\omega} \cdot\mathbf{S} + V(|\mathbf{r}|) \tag*{(19)}

We can rewrite this equation as

i∂∂t=−12μ∇2−ωLz−ωSz+V(∣r∣)(20)i\frac{\partial}{\partial t} = -\frac{1}{2\mu}\nabla^2 - \omega L_z - \omega S_z + V(|\mathbf{r}|) \tag*{(20)}

where we used p=−iℏ∇\mathbf{p} = -i\hbar\nabla, H=i∂∂t\mathcal{H} = i\frac{\partial}{\partial t}, ω=ωz^\boldsymbol{\omega} = \omega\hat{\mathbf{z}} and Lz=−i∂∂ϕL_z = -i\frac{\partial}{\partial\phi}.

In rotating frames, the contribution in Hamiltonian is only the product between orbital angular momentum due to the vortices with the spin of meson. Thus, in this non-inertial frame, the contribution is just to LL different from zero. Then, we will write just the radial part of the equation 21.

1r2ddr(r2dR(r)dr)+[2μ(E−V(r))−l(l+1)r2−2μωLz−2μωSz]R(r)=0(21)\frac{1}{r^2}\frac{d}{dr}\left(r^2\frac{dR(r)}{dr}\right) + \left[2\mu(E-V(r))-\frac{l(l+1)}{r^2}-2\mu\omega L_z-2\mu\omega S_z\right]R(r)=0 \tag*{(21)}

To go forward, we shall assume the rotation to be classical and related to the hydrodynamic vorticity. Then we can define ω\omega in terms of the a conserved circulation

C=∮v⋅dl(22)C = \oint\mathbf{v}\cdot d\mathbf{l} \tag*{(22)}

and also assume the Cornell potential [22]

V(r)=br−αeffr(23)V(r) = b r - \frac{\alpha_{eff}}{r} \tag*{(23)}

so

12μr2ddr(r2dR(r)dr)+[(E−V(r))−l(l+1)2μr2−mjCπr2]R(r)=0(24)\frac{1}{2\mu r^2}\frac{d}{dr}\left(r^2\frac{dR(r)}{dr}\right) + \left[(E - V(r)) - \frac{l(l+1)}{2\mu r^2} - \frac{m_j C}{\pi r^2}\right]R(r) = 0 \tag*{(24)}

Making χ(r)=rR(r)\chi(r) = rR(r) and change the variable x=1/rx = 1/r, then the equation 25, we get:

d2χ(x)dx2+2xx2dχ(x)dx+2μx4[E−bx+αeffx−l(l+1)2μx2−mjC2πx2]χ(x)=0(25)\frac{d^2\chi(x)}{dx^2} + \frac{2x}{x^2}\frac{d\chi(x)}{dx} + \frac{2\mu}{x^4}\left[E - \frac{b}{x} + \alpha_{eff}x - \frac{l(l+1)}{2\mu}x^2 - \frac{m_j C}{2\pi}x^2\right]\chi(x) = 0 \tag*{(25)}

Now we will expand the variable yy (y=x−δy = x - \delta) around zero with δ=1/r0\delta= 1/r_0 where r0r_0 is the mean meson radius. Therefore we have:

bx≈b(3δ−3xδ2+x2δ3)(26)\frac{b}{x} \approx b\left(\frac{3}{\delta} - \frac{3x}{\delta^2} + \frac{x^2}{\delta^3}\right) \tag*{(26)}

Thus we can write the equation 26 in following form:

d2χ(x)dx2+2xx2dχ(x)dx+2μx4[E−3bδ+(αeff+3bδ2)x−(l(l+1)2μ+mjC2π+bδ3)x2]χ(x)=0(27)\frac{d^2\chi(x)}{dx^2} + \frac{2x}{x^2}\frac{d\chi(x)}{dx} + \frac{2\mu}{x^4}\left[E - \frac{3b}{\delta} + \left(\alpha_{eff} + \frac{3b}{\delta^2}\right)x - \left(\frac{l(l+1)}{2\mu} + \frac{m_j C}{2\pi} + \frac{b}{\delta^3}\right)x^2\right]\chi(x) = 0 \tag*{(27)}

The coefficients HiH_i with i=0,1,2i = 0, 1, 2, are given by:

H0=−2μ(E−3bδ)(28)H_0 = -2\mu\left(E - \frac{3b}{\delta}\right) \tag*{(28)}
H1=2μ(αeff+3bδ2)(29)H_1 = 2\mu\left(\alpha_{eff} + \frac{3b}{\delta^2}\right) \tag*{(29)}
H2=−2μ(l(l+1)2μ+mjC2π+bδ3)(30)H_2 = -2\mu\left(\frac{l(l+1)}{2\mu} + \frac{m_j C}{2\pi} + \frac{b}{\delta^3}\right) \tag*{(30)}

Thus, we can write:

d2χ(x)dx2+2xx2dχ(x)dx+2μx4[−H0+H1x+H2x2]R(x)=0(31)\frac{d^2\chi(x)}{dx^2} + \frac{2x}{x^2}\frac{d\chi(x)}{dx} + \frac{2\mu}{x^4}\left[-H_0 + H_1 x + H_2 x^2\right]R(x) = 0 \tag*{(31)}

Now, we can compare the differential equation 32 with A1, we get:

σ(x)=x2 ,τ~(x)=2x ,σ~(x)=−H0+H1x+H2x2(32)\sigma(x) = x^2 \,, \qquad\widetilde{\tau}(x) = 2x \,, \qquad\widetilde{\sigma}(x) = -H_0 + H_1 x + H_2 x^2 \tag*{(32)}

The function π(s)\pi(s) from equation (60) is given by:

π(x)=±(k−H2)x2−H1x+H0(33)\pi(x)=\pm\sqrt{(k-H_{2})x^{2}-H_{1}x+H_{0}} \tag*{(33)}

Choosing the negative solution the polynomial inside of square must have discriminant equal to zero, so we get:

k=14H0(H12+4H0H2)(34)k=\frac{1}{4H_{0}}\left(H_{1}^{2}+4H_{0}H_{2}\right) \tag*{(34)}

Then,

π(x)=−12H0(H1x−2H0)(35)\pi(x)=-\frac{1}{2\sqrt{H_{0}}}\left(H_{1}x-2H_{0}\right) \tag*{(35)}

and τ(x)\tau(x) is given by equation (63), so

τ(x)=2x−1H0(H1x−2H0)(36)\tau(x)=2x-\frac{1}{\sqrt{H_{0}}}\left(H_{1}x-2H_{0}\right) \tag*{(36)}

Then equaling the equations A6 and A7 we have:

H124H0+H2−H12H0=H1H0n−n(n+1)(37)\frac{H_{1}^{2}}{4H_{0}}+H_{2}-\frac{H_{1}}{2\sqrt{H_{0}}}=\frac{H_{1}}{\sqrt{H_{0}}}n-n(n+1) \tag*{(37)}

Solving the equation 38 to H0H_{0}, we get:

H0=H1(1+2n)±1−4H2(38)\sqrt{H_{0}}=\frac{H_{1}}{(1+2n)\pm\sqrt{1-4H_{2}}} \tag*{(38)}

In this way, we can obtain the energy levels and wavefunctions as a function of the rotation parameters via the Eigenvalue equation

−2μ(E−3bδ)=2μ(αeff+3bδ2)(1+2n)+1+4l(l+1)+4μmjCπ+8μbδ3(39)\sqrt{-2\mu\left(E-\frac{3b}{\delta}\right)}= \frac{2\mu\left(\alpha_{eff}+\frac{3b}{\delta^{2}}\right)} {(1+2n)+\sqrt{1+4l(l+1)+\frac{4\mu m_{j}C}{\pi}+\frac{8\mu b}{\delta^{3}}}} \tag*{(39)}

Mass and vorticity

The previous equation gives the binding energy in the rotating frame

En,l,m=3bδ−2μ(αeff+3bδ2)2[(1+2n)+1+4l(l+1)+4μmjCπ+8μbδ3]2(40)E_{n,l,m}=\frac{3b}{\delta}- \frac{2\mu\left(\alpha_{eff}+\frac{3b}{\delta^{2}}\right)^{2}} {\left[(1+2n)+\sqrt{1+4l(l+1)+\frac{4\mu m_{j}C}{\pi}+\frac{8\mu b}{\delta^{3}}}\right]^{2}} \tag*{(40)}

With he wave function determination is explicitly given in Appendix § 1.0.3 using the method outlined in Appendix § 1.0.1. We can obtain the quarkonium mass from the equation (40):

M=2mq+En,l,m(41)M=2m_{q}+E_{n,l,m} \tag*{(41)}
M=2mq+3bδ−2μ(αeff+3bδ2)2[(1+2n)+1+4l(l+1)+4μmjCπ+8μbδ3]2(42)M = 2m_q + \frac{3b}{\delta} - \frac{2\mu\left(\alpha_{eff}+\frac{3b}{\delta^2}\right)^2}{\left[(1+2n)+\sqrt{1+4l(l+1)+\frac{4\mu m_j C}{\pi}+\frac{8\mu b}{\delta^3}}\right]^2} \tag*{(42)}

Figure 4 and Figure 5 gives a quantitative value of the binding energy for quarkonia polarized with and opposite the vorticity as a function of the circulation. Unsurprisingly, one is the opposite of the other. This, however, is not experimentally detectable since spin alignment measurements do not distinguish between m=±1m=\pm1 states. The mass difference in section 4 would therefore appear as an impact parameter dependent widening of the quarkonium state. However, since a similar widening occurs in any kind of in-medium interaction, particularly in interactions leading to the melting of quarkonium [5], a univocal proof of spin-orbit coupling can not be obtained by measuring E±1E_{\pm1} alone. As we will show in the next section section IV, however, off-diagonal matrix elements could be of help here.

Figure 5. Bottomonium mass in relation to parameter CC for various energy and momentum levels. Notation follows atomic physics quantum numbers using T=168T = 168 MeV

Vorticity and melting

We can use the mass dependence calculated in the previous sub-section to study, qualitatively, how the temperature for quarkonium melting (melting temperature, TmeltT_{\rm melt}) changes under a non-inertial frame. This calculation can be considered to be a semi-classical estimate of the imaginary part of the energy calculated in the previous subsection section III B.

Using a semi-classical analysis with ⟨p⟩∼1/r\langle p\rangle\sim1/r and ⟨p2⟩∼1/r2\langle p^2\rangle\sim1/r^2 we can write the energy by using

E(r)=12μr2−mjω+V(r)(43)E(r)=\frac{1}{2\mu r^2}-m_j\omega+V(r) \tag*{(43)}

For the J/ψJ/\psi melting, we shall use the potential that takes into consideration the Debye screen where λD\lambda_D is Debye screening length, which is given by:

V(r)=−αeffe−r/λDr(44)V(r)=-\alpha_{eff}\frac{e^{-r/\lambda_D}}{r} \tag*{(44)}

Next, using the circulation theorem 23. Then,

E(r)=[12μ−mjC2π]1r2−αeffe−r/λDr(45)E(r)=\left[\frac{1}{2\mu}-\frac{m_j C}{2\pi}\right]\frac{1}{r^2}-\frac{\alpha_{eff}e^{-r/\lambda_D}}{r} \tag*{(45)}

The bound state is defined when the energy 46 has a minimum, so we can write:

dE(r)dr=−[1μ−mjCπ]1r3+αeffe−r/λD[1+rλD]1r2=0(46)\frac{dE(r)}{dr}=-\left[\frac{1}{\mu}-\frac{m_j C}{\pi}\right]\frac{1}{r^3}+\alpha_{eff}e^{-r/\lambda_D}\left[1+\frac{r}{\lambda_D}\right]\frac{1}{r^2}=0 \tag*{(46)}

, then

r[1+rλD]αeffe−r/λD=1μ−mjCπ(47)r\left[1+\frac{r}{\lambda_D}\right]\alpha_{eff}e^{-r/\lambda_D}=\frac{1}{\mu}-\frac{m_j C}{\pi} \tag*{(47)}

Making the following variable change r~=r/λD\tilde{r}=r/\lambda_D, we get:

f(r~)=r~(1+r~)e−r~=1αeffλD[1μ−mjCπ](48)f(\tilde{r})=\tilde{r}(1+\tilde{r})e^{-\tilde{r}}=\frac{1}{\alpha_{eff}\lambda_D}\left[\frac{1}{\mu}-\frac{m_j C}{\pi}\right] \tag*{(48)}

The maximum value of f(r~)f(\tilde{r}) is 0.840 at r~=1.92\tilde{r}=1.92 in this limit we will have the non bound states. Thereby, we can write the inequality for this limit:

1αeffλD[1μ−mjCπ]>0.840(49)\frac{1}{\alpha_{eff}\lambda_D}\left[\frac{1}{\mu}-\frac{m_j C}{\pi}\right]>0.840 \tag*{(49)}

At this point, we will utilize the Debye lenght λD\lambda_D value from the lowest-order perturbative QCD [31].2

λD=29παeff1T(50)\lambda_D=\sqrt{\frac{2}{9\pi\alpha_{eff}}\frac{1}{T}} \tag*{(50)}

We can obtain an estimate for the melting temperature TmeltT_{melt} in non-inertial frames using the equation 49 in the f(r~)f(\tilde{r}) maximum value, then:

0.84029παeff1Tmelt=1αeff[1μ−mjCπ](51)0.840\sqrt{\frac{2}{9\pi\alpha_{eff}}\frac{1}{T_{melt}}}=\frac{1}{\alpha_{eff}}\left[\frac{1}{\mu}-\frac{m_j C}{\pi}\right] \tag*{(51)}

So,

Tmelt=0.8402αeff9π[1μ−mjCπ](52)T_{melt}=0.840\sqrt{\frac{2\alpha_{eff}}{9\pi}}\left[\frac{1}{\mu}-\frac{m_j C}{\pi}\right] \tag*{(52)}

We can see from figure 6 that the melting temperature TmeltT_{melt} depends on the spin polarization quantum number mm. In particular for m=1m=1 this temperature increases considerably. For m=0m=0 it predictably does not change and for m=−1m=-1 it decreases but, and this is a fundamental point, it decreases a lot less than the increase for m=1m=1. Therefore, in a vortical medium polarized quarkonia will be much more likely to survive, while un-polarized or anti-polarized quarkonium’s survival probability does not change that much w.r.t. quarkonia in a non-rotating medium.

This finding is qualitatively important since it shows that vorticity can link quarkonium suppression and polarization via ”distillation”. In a vortical medium, the melting probability of quarkonium states will depend strongly on their polarization. Such a mechanism will result, analogously to the η\eta [33], to a strong and novel dependence of quarkonium abundance on centrality, which could be investigated quantitatively by a hydrodynamic model.

One can be ”brave” and try to apply our potential model to the meson ϕ\phi, which is formed of strange and anti-strange quarks. The strong-vorticity dependent melting might be able to explain the strong spin alignment observed in experiment, which seems incompatible with Cooper-Frye freezeout [8]. In the picture described here, melting temperature for aligned ϕ\phi rises a lot, while melting temperature for anti-aligned and non-aligned ϕ\phi stays nearly the same. Thus, the large apparent spin alignment of ϕ\phis comes from a ”distillation” process where only polarized ϕ\phis survive, and this increases relative ϕ\phi spin alignment. Of course, applying potential models to ϕ\phi is not justified theoretically and requires model-building, so such a solution will need considerable quantitative and phenomenological development.

Figure 6. An estimate of melting temperature TmeltT_{\mathrm{melt}} for J/ψJ/\psi (left) and Υ\Upsilon (right) melting to different spin polarization mm values.

In the absence of a detailed hydrodynamic modeling of the correlation between charmonium suppression and polarization, however, distinguishing the effects outlined above from other variations of charmonium suppression with centrality, namely the effect of centrality dependence of temperature, looks complicated, with vorticity just adding an “event-by-event widening” to the processes of melting and regeneration of quarkonia. The next section, however, provides a direct experimentally measurable indication of spin-orbit non-equilibrium in quarkonium polarization measurements

Density matrix elements and vorticity

Turning our attention back to the density matrix we can write this operator on basis of energy in the following way:

ρ^=e−βH^,β=1T(53)\hat{\rho}=e^{-\beta\hat{H}}, \qquad\beta=\frac{1}{T} \tag*{(53)}

Where the eigenvalues is given by the equation (40). In this moment, we will make a rotation to lab frame, so:

ρ^r=U(θr,ϕr)ρ^U−1(θr,ϕr)(54)\hat{\rho}^{r}=U(\theta_{r},\phi_{r})\hat{\rho}U^{-1}(\theta_{r},\phi_{r}) \tag*{(54)}

Now, we can expand the equation 55 in following way:

ρm,m′r=∑m′′,m′′′ei(m′′′−m)ϕrdm,m′′j(θr)ρm′′,m′′′[dm′′′,m′j(θr)]−1,ρm′′,m′′′=1ZeβEm′′′δm′′,m′′′(55)\rho^{r}_{m,m'}=\sum_{m'',m'''}e^{i(m'''-m)\phi_{r}}d^{j}_{m,m''}(\theta_{r})\rho_{m'',m'''}\left[d^{j}_{m''',m'}(\theta_{r})\right]^{-1}, \qquad\rho_{m'',m'''}=\frac{1}{Z}e^{\beta E_{m'''}}\delta_{m'',m'''} \tag*{(55)}

We can relate the density matrix coefficients from variable CC as both density matrix coefficients and energy variation depend on the parameter CC. So we can relate these two values we obtain the Figure 8. Then, we can relate the energy (ΔEm,m′=Em−Em′)(\Delta E_{m,m'}=E_m-E_{m'}) with off-diagonal density matrix Re[ρ−1,0]Re[\rho_{-1,0}] and Re[ρ−1,1]Re[\rho_{-1,1}] using the values ϕ=0\phi=0, θr=1.75±0.10\theta_{r}=1.75\pm0.10 to Collins-Soper refer to transverse momentum range 2<pT<42<p_{T}<4 (GeV).

Figure 8. The off-diagonal density matrix components in relation to energy variation ΔEm,m′\Delta E_{m,m'} using T=168T = 168 MeV.

As can be seen in Figure 7 the obtained ρ00\rho_{00} are compatible with the estimates shown in section II

Figure 7. The alignment factor ρ00\rho_{00} in relation C for charmonium and bottomonium using T=168T = 168 MeV.

We can obtain the off-diagonal density matrix components in relation to the quarkonium energy using the parameters shown in Table II. In Figure 8 and 9, we can note the relation between the alignment factor, ρ00\rho_{00}, with the circulation parameter CC. It is evident that an

mm (GeV)αeff\alpha_{eff}bb (GeV2^{2})δ\delta (GeV)μ\mu (GeV)θr\theta_{r} (rad)
cˉcˉ\bar{c}\bar{c}1.2091.2440.20.2310.60451.75 ±\pm 0.10
bˉbˉ\bar{b}\bar{b}4.8231.5690.20.3782.41151.58 ±\pm 0.10

Table II. TABLE II. Eigenvalue quarkonium parameters obtained from [30] with θr\theta_{r} determined in system equation 9 to Collins-Soper frame to J/ψJ/\psi and Υ\Upsilon with respectively ranges 4<pT<64<p_{T}<6 (GeV)(GeV) and pT<15p_{T}<15 (GeV)(GeV).

increase of C increments ρ00\rho_{00}; however, this increase will depend on the type of meson. As we can see the bottomonium alignment factor is larger than the charmonium when compared under the same parameter C.

Unlike the figures in the previous sections, the x,yx, y axes of the plots in this section are independently measurable. ΔE\Delta E will be manifest event-by-event as an invariant mass correlation. ρi,j\rho_{i,j} can be obtained [9] from dN/dΩdN/d\Omega according to a harmonic analysis of Eq. (1). Barring acceptance effects, there is no non-vortical dynamics capable of generating a correlation between these two quantities. The most direct analysis to do is a correlation between λθ,ϕ,θϕ\lambda_{\theta,\phi,\theta\phi} and the invariant mass and width. A definite signal of the correlation can be used, using the functions described in this section, to extract the angle between vorticity and heavy quark spin polarization. This can then be used to constrain models of spin hydrodynamics such as [13, 14, 15, 16, 17, 18, 19, 20]. This section’s results, therefore, can be used as a baseline for developing an experimental analysis capable of probing non-equilibrium between spin and vorticity using quarkonium probes.

The most direct analysis to do is a correlation between λθ,ϕ,θϕ\lambda_{\theta,\phi,\theta\phi} and the invariant mass and width. A definite signal of the correlation can be used, using the functions described in this section, to extract the angle between vorticity and heavy quark spin polarization. This can then be used to constrain models of spin hydrodynamics such as [13, 14, 15, 16, 17, 18, 19, 20]. This section’s results, therefore, can be used as a baseline for developing an experimental analysis capable of probing non-equilibrium between spin and vorticity using quarkonium probes.

Figure 9. The off-diagonal density matrix components in relation to energy variation ΔEm,m′\Delta E_{m,m'} using T=168T = 168 MeV.

Conclusion

The results obtained in this work are certainly to be considered more like a rough estimate than a quantitative analysis, since they use an early version of the Cornell potential model. However we think that this model is good enough to get a physical intuition of the problem of linking spin-vorticity non-equilibrium to the quarkonium state in a rotating frame, and, respectively, the quarkonium state to experimental observables. As § 1 illustrates, our formalism in fact reproduces the reasonable expectation of how the binding energy (mass) and melting probability (width) of the quarkonium state respond to rotation.

Section IV therefore takes the model further and examines what happens when rotation and spin are not aligned, and suggests experimental measurements. A positive experimental observation of the observables described in the previous section will provide evidence that quarkonium state polarization is not thermally aligned, but in fact “remembers” a polarization state which is quite distinct from the vorticity state.

As argued in section II and Figure 1, one in fact expects quarks to be aligned transversely, which would also follow from initial state dynamics [34, 35] and from the large transverse vorticity present in the beginning of the collision [6]. However, the vortex where the quarkonia will be could up could well be aligned to the flow gradients that developed on the time-scale of a hydrodynamic expansion, which will be in the longitudinal direction. The combination of the two will result in a density matrix exhibiting large corrections to the expectation value in [32], with the magnitude of these correcting correlated to the invariant mass of quarkonium.

A very interesting associated result is the strong dependence, non-monotonic in both TT and mm, of the dissociation temperature TmeltT_{\mathrm{melt}} of quarkonium shown in § 1 C. It suggests a new ”distillation” mechanism for quarkonium polarization, where aligned quarkonia have a stronger probability to survive in medium while non-aligned and anti-aligned quarkonia simply melt. Speculatively considering the ϕ\phi a quarkonium state could explain the strong alignment signal seen in [8], though some phenomenological work is needed to confirm such a model is viable.

So far, as seen in section II there is no direct experimental evidence for such non-equilibrium. In the short term, an azimuthal modulation of λθ,ϕ,θϕ\lambda_{\theta,\phi,\theta\phi} (Figure 3) could provide a strong indication that such non-equilibrium should be investigated. In the long term, an experiment with sufficient resolution in invariant mass and capable of reconstructing the spin alignment could be able to perform the analysis advocated in section IV, providing more direct experimental evidence.

Quarkonium polarization, therefore, could well be a very promising observable for experimentally probing spin hydrodynamics in the quark-gluon plasma, a theoretically interesting and challenging topic that so far had little contact with phenomenology.

GT thanks CNPQ bolsa de produtividade 306152/2020-7, bolsa FAPESP 2021/01700-2 and participation in tematic FAPESP, 2017/05685-2 and grant BPN/ULM/2021/1/00039 from the Polish National Agency for Academic Exchange K.J.G. is supported by CAPES doctoral fellowship 88887.464061/2019-00. This work was supported in part by the Polish National Science Centre Grant Nos 2018/30/E/ST2/00432. We thank the hospitality of the Jagellonian university when part of this work was performed.

Nikiforov-Uvarov method

We will use the method developed by Nikiforov and Uvarov to solve a differential equation where it is possible to reduce to hypergeometric function. We can do it too to Schrödinger equation, which can write in the following form:

ψ′′(s)+τ~(s)σ(s)ψ′(s)+σ~(s)σ2(s)ψ(s)=0(56)\psi^{\prime\prime}(s) + \frac{\widetilde{\tau}(s)}{\sigma(s)}\psi^{\prime}(s) + \frac{\widetilde{\sigma}(s)}{\sigma^{2}(s)}\psi(s) = 0 \tag*{(56)}

Using the variable separation technique,

ψ(s)=ϕ(s)y(s)(57)\psi(s) = \phi(s)y(s) \tag*{(57)}

Using the variable separation technique, we can see that the equation must satisfy the following expression [26]:

ϕ′(s)ϕ(s)=π(s)σ(s)(58)\frac{\phi^{\prime}(s)}{\phi(s)} = \frac{\pi(s)}{\sigma(s)} \tag*{(58)}

so we can simplify as

σ(s)y′′(s)+τ(s)y′(s)+λy(s)=0(59)\sigma(s)y^{\prime\prime}(s) + \tau(s)y^{\prime}(s) + \lambda y(s) = 0 \tag*{(59)}

At this moment, we need to define the π(s)\pi(s) function in the following form:

π(s)=σ′(s)−τ~(s)2±(σ′(s)−τ~(s)2)2−σ~(s)+kσ(s)(60)\pi(s) = \frac{\sigma^{\prime}(s) - \widetilde{\tau}(s)}{2} \pm\sqrt{\left(\frac{\sigma^{\prime}(s) - \widetilde{\tau}(s)}{2}\right)^{2} - \widetilde{\sigma}(s) + k\sigma(s)} \tag*{(60)}

and λ\lambda parameter is given by:

λ=k+π′(s)(61)\lambda= k + \pi^{\prime}(s) \tag*{(61)}

The Schrödinger equation eigenvalue obtained from this method is given by:

λn=−nτ′(s)−n(n−1)2σ′′(s)(62)\lambda_n=-n\tau'(s)-\frac{n(n-1)}{2}\sigma''(s) \tag*{(62)}

where τ(s)\tau(s) is defined by:

τ(s)=τ~(s)+2π(s)(63)\tau(s)=\widetilde{\tau}(s)+2\pi(s) \tag*{(63)}

The solution to hypergeometric equation is given by Rodrigues relation:

yn(s)=Bnρ~(s)dndsn[σnρ~(s)](64)y_n(s)=\frac{B_n}{\widetilde{\rho}(s)}\frac{d^n}{ds^n}\left[\sigma^n\widetilde{\rho}(s)\right] \tag*{(64)}

BnB_n is the normalization coefficient and ρ~(s)\widetilde{\rho}(s) must obey the following relation:

(σ(s)ρ~(s))′=τ(s)ρ~(s)(65)\left(\sigma(s)\widetilde{\rho}(s)\right)'=\tau(s)\widetilde{\rho}(s) \tag*{(65)}

Then

ρ~′ρ~=τ−σ′σ(66)\frac{\widetilde{\rho}'}{\widetilde{\rho}}=\frac{\tau-\sigma'}{\sigma} \tag*{(66)}

We have the following solution:

ρ~(s)=exp⁡(∫τ−σ′σ ds)(67)\widetilde{\rho}(s)=\exp\left(\int\frac{\tau-\sigma'}{\sigma}\,ds\right) \tag*{(67)}

Estimate of λθ,ϕ,θϕ\lambda_{\theta,\phi,\theta\phi}

The blast wave model, described in detail in [38], is to parametrize the transverse plane flow profile of the fireball to obtain soft observables and relate them to particle properties via a Cooper-Frye type formula [32]. The fireball is assumed to be an ellipse in coordinate space

x=rmax⁡1−ϵcos⁡ϕ,y=rmax⁡1+ϵsin⁡ϕ,(68)x=r_{\max}\sqrt{1-\epsilon}\cos\phi,\qquad y=r_{\max}\sqrt{1+\epsilon}\sin\phi, \tag*{(68)}

where the rmax⁡r_{\max} and ϵ\epsilon represents the parameters of model and ϕ\phi is the azimutal angle in the transverse plane. To obtain the thermal vorticity, we need first define the fluid velocity that describes the azimuthally asymmetric fluid flow, parametrized by a Hubble-type Ansatz

uμ=(tN,xN1+δ~,yN1−δ~,zN)(69)u^\mu=\left(\frac{t}{N},\frac{x}{N}\sqrt{1+\widetilde{\delta}},\frac{y}{N}\sqrt{1-\widetilde{\delta}},\frac{z}{N}\right) \tag*{(69)}

Where the normalization factor is equal to N=τ2−(x2−y2)δ~N = \sqrt{\tau^2 - (x^2-y^2)\tilde{\delta}} and the proper time is τ2=t2−x2−y2−z2\tau^2 = t^2 - x^2 - y^2 - z^2 ensuring, uμuμ=1u^\mu u_\mu= 1. At this point, we can obtain the thermal vorticity from the equation eq (5) defined above:

ωμν=−12T(∂μuν−∂νuμ)−12T2(uμ∂νT−uν∂μT).(70)\omega_{\mu\nu} = -\frac{1}{2T}(\partial_\mu u_\nu-\partial_\nu u_\mu)-\frac{1}{2T^2}(u_\mu\partial_\nu T-u_\nu\partial_\mu T). \tag*{(70)}

The components are given explicitly in [38]. Now, we will suppose that the ω\omega is defined in the following way:

ω=ωx2+ωy2+ωz2(71)\omega= \sqrt{\omega_x^2+\omega_y^2+\omega_z^2} \tag*{(71)}

Where the spacial components to vorticity are given by ωi=12ϵijkωjk\omega^i = \frac{1}{2}\epsilon^{ijk}\omega_{jk}. Now, we will make the average on the azimuthal angle, ϕ\phi, using the parameters to centrality 30–60% in midrapidity η=0\eta= 0 in equation B4. We have:

⟨ω⟩=12π∫02πω(rmax,ϵ,δ~)dϕ=24.08 MeV(72)\langle\omega\rangle= \frac{1}{2\pi}\int_0^{2\pi}\omega(r_{max},\epsilon,\tilde{\delta})d\phi= 24.08\ MeV \tag*{(72)}

We can make an estimative for this value for thermal vorticity B5 using equation 23 in the following form: C=2π⟨ω⟩r02C = 2\pi\langle\omega\rangle r_0^2. Now, using r0=1/δr_0 = 1/\delta, we have:

C=2π⟨ω⟩δ2(73)C = \frac{2\pi\langle\omega\rangle}{\delta^2} \tag*{(73)}

Using the numerical parameters reported in [38, 39] via a fit to experimental data, we can obtain the vorticity.

To obtain an estimate of the maximum value of λθ,ϕθϕ\lambda_{\theta,\phi\theta\phi} we also need a numerical estimate of θr,ϕr\theta_r,\phi_r in Eq. (4). This can be obtained from inverting Figure 2. The final result, calculated by plugging in these values in equations 2, B6 and 56 is shown in Table I.

This is a very rough estimate, since these parameters were fitted at a different energy, and since only the second Fourier coefficient of the flow is considered (in vorticity all coefficients mix non-trivially). The latter reason means we cannot calculate the exact azimuthal dependence. However, since vnv_n changes slowly with energy we are confident this is a good order-of-magnitude estimate of the effect experimentalists are looking for. If azimuthal modulation is found, more realistic models will be needed to describe it quantitatively.

Quarkonium wave function determination

Now, we can obtain the wave function from the equation A12 and using the equation 37 and σ(x)=x2\sigma(x) = x^2. From it, we get

ρ~(x)=C1x−H1/H0e−2H0/x(74)\tilde{\rho}(x) = C_1 x^{-H_1/\sqrt{H_0}}e^{-2\sqrt{H_0}/x} \tag*{(74)}

We are able to obtain the value of function ϕ(x)\phi(x) from equation (35) and using the values of equation (35). Then

ϕ′(x)ϕ(x)=−12H0(H1x−2H0x2)(75)\frac{\phi'(x)}{\phi(x)}=-\frac{1}{2\sqrt{H_0}}\left(\frac{H_1}{x}-\frac{2H_0}{x^2}\right) \tag*{(75)}

Then

ϕ(x)=C2x−H1/(2H0)e−H0/x(76)\phi(x)=C_2x^{-H_1/(2\sqrt{H_0})}e^{-\sqrt{H_0}/x} \tag*{(76)}

We can determine the yn(r)y_n(r) from equations (64), (74) and σ(r)\sigma(r). Then,

yn(r)=BnxH1/H0e2H0/xdndxn[x2n−H1/H0e−2H0/x](77)y_n(r)=B_nx^{H_1/\sqrt{H_0}}e^{2\sqrt{H_0}/x}\frac{d^n}{dx^n}\left[x^{2n-H_1/\sqrt{H_0}}e^{-2\sqrt{H_0}/x}\right] \tag*{(77)}

From equation (76), C4 and putting in A2, we have that

ψ(x)=BnxH1/(2H0)eH0/xdndxn[x2n−H1/H0e−2H0/x](78)\psi(x)=B_nx^{H_1/(2\sqrt{H_0})}e^{\sqrt{H_0}/x}\frac{d^n}{dx^n}\left[x^{2n-H_1/\sqrt{H_0}}e^{-2\sqrt{H_0}/x}\right] \tag*{(78)}

Thereby,changing the variable x=1/rx=1/r we can write the radial wave fuction from equation (78) and relation ψ(r)=rR(r)\psi(r)=rR(r). So:

Rn(r)=Bnr−(H1/2H0)eH0r[−r2ddr]n[x2n−H1/H0e−2H0/x](79)R_n(r)=B_nr^{-(H_1/2\sqrt{H_0})}e^{\sqrt{H_0}r}\left[-r^2\frac{d}{dr}\right]^n\left[x^{2n-H_1/\sqrt{H_0}}e^{-2\sqrt{H_0}/x}\right] \tag*{(79)}

Thus, we can write the wave fuction that following form:

Ψn,l,m(r,θ,ϕ)=Rn(r)Ylm(θ,ϕ)(80)\Psi_{n,l,m}(r,\theta,\phi)=R_n(r)Y_l^m(\theta,\phi) \tag*{(80)}

Using the value of C6, we get:

Ψn,l,m(r,θ,ϕ)=Bnr−(H1/2H0)eH0r[−r2ddr]n[x2n−H1/H0e−2H0/x]Ylm(θ,ϕ)(81)\Psi_{n,l,m}(r,\theta,\phi)=B_nr^{-(H_1/2\sqrt{H_0})}e^{\sqrt{H_0}r}\left[-r^2\frac{d}{dr}\right]^n\left[x^{2n-H_1/\sqrt{H_0}}e^{-2\sqrt{H_0}/x}\right]Y_l^m(\theta,\phi) \tag*{(81)}

References

References

  1. [4]R. L. Thews, M. Schroedter and J. Rafelski, Phys. Rev. C 63 (2001), 054905 doi:10.1103/PhysRevC.63.054905 [arXiv:hep-ph/0007323 [hep-ph]].DOI
  2. [5]A. Rothkopf, Phys. Rept. 858 (2020), 1-117 doi:10.1016/j.physrep.2020.02.006 [arXiv:1912.02253 [hep-ph]].DOI
  3. [6]F. Becattini and M. A. Lisa, Ann. Rev. Nucl. Part. Sci. 70 (2020), 395-423 doi:10.1146/annurev-nucl-021920-095245 [arXiv:2003.03640 [nucl-ex]].DOI
  4. [7]S. Acharya et al. [ALICE], [arXiv:1910.14408 [nucl-ex]].
  5. [8]M. S. Abdallah et al. [STAR], Nature 614, no.7947, 244-248 (2023) doi:10.1038/s41586-022-05557-5 [arXiv:2204.02302 [hep-ph]].DOI
  6. [9]Z. T. Liang and X. N. Wang, Phys. Lett. B 629, 20-26 (2005) doi:10.1016/j.physletb.2005.09.060 [arXiv:nucl-th/0411101 [nucl-th]].DOI
  7. [10]X. L. Xia, H. Li, X. G. Huang and H. Zhong Huang, Phys. Lett. B 817 (2021), 136325 doi:10.1016/j.physletb.2021.136325 [arXiv:2010.01474 [nucl-th]].DOI
  8. [11]K. J. Gonçalves and G. Torrieri, Phys. Rev. C 105 (2022) no.3, 034913 doi:10.1103/PhysRevC.105.034913 [arXiv:2104.12941 [nucl-th]].DOI
  9. [12]X. L. Sheng, L. Oliva, Z. T. Liang, Q. Wang and X. N. Wang, [arXiv:2205.15689 [nucl-th]].DOI
  10. [13]D. Montenegro and G. Torrieri, Phys. Rev. D 100, no.5, 056011 (2019) doi:10.1103/PhysRevD.100.056011 [arXiv:1807.02796 [hep-th]].DOI
  11. [14]D. Montenegro and G. Torrieri, [arXiv:2004.10195 [hep-th]].
  12. [15]D. Montenegro and G. Torrieri, Phys. Rev. D 107 (2023) no.7, 076010 doi:10.1103/PhysRevD.107.076010 [arXiv:2207.00537 [hep-th]].DOI
  13. [16]S. Bhadury, W. Florkowski, A. Jaiswal, A. Kumar and R. Ryblewski, Phys. Lett. B 814 (2021), 136096 doi:10.1016/j.physletb.2021.136096 [arXiv:2002.03937 [hep-ph]].DOI
  14. [17]S. Shi, C. Gale and S. Jeon, [arXiv:2008.08618 [nucl-th]].
  15. [18]A. D. Gallegos, U. Gürsoy and A. Yarom, SciPost Phys. 11 (2021), 041 doi:10.21468/SciPostPhys.11.2.041 [arXiv:2101.04759 [hep-th]].DOI
  16. [19]M. Hongo, X. G. Huang, M. Kaminski, M. Stephanov and H. U. Yee, JHEP 11 (2021), 150 doi:10.1007/JHEP11(2021)150 [arXiv:2107.14231 [hep-th]].DOI
  17. [20]K. Hattori, M. Hongo, X. G. Huang, M. Matsuo and H. Taya, Phys. Lett. B 795 (2019), 100-106 doi:10.1016/j.physletb.2019.05.040 [arXiv:1901.06615 [hep-th]].DOI
  18. [21]Acharya S, Adamová D, Adler A, Adolfsson J, Aggarwal MM, Aglieri Rinella G, et al. First measurement of quarkonium polarization in nuclear collisions at the LHC. Physics Letters B. 2021;815:136146.
  19. [22]N. Brambilla, A. Pineda, J. Soto and A. Vairo, Nucl. Phys. B 566 (2000), 275 doi:10.1016/S0550-3213(99)00693-8 [arXiv:hep-ph/9907240 [hep-ph]].DOI
  20. [23]Lucha, W. (1991). Bound states of Quarks. Physics Reports, 200(4), 127-240. https://doi.org/10.1016/0370-1573(91)90001-3
  21. [24]M. Buzzegoli and K. Tuchin, Nucl. Phys. A 1030 (2023), 122577 doi:10.1016/j.nuclphysa.2022.122577 [arXiv:2209.03991 [nucl-th]].DOI
  22. [25]H. Kim, S. Cho and S. H. Lee, [arXiv:2212.14570 [hep-ph]].DOI
  23. [26]Yazdankish, E. (2020). Solving of the Schrodinger equation analytically with an approximated scheme of the woods-saxon potential by the systematical method of Nikiforov-Uvarov. International Journal of Modern Physics E, 29(06), 2050032. https://doi.org/10.1142/S0218301320500329
  24. [27]Arfken, G. B. (2005). Mathematical Methods for Physicists (3rd ed.). Elsevier Academic Press.
  25. [28]Anandan, J., Suzuki, J. (n.d.). Quantum Mechanics in a Rotating Frame. https://arxiv.org/pdf/quant-ph/0305081.pdf
  26. [29]Anandan, J. (1992). Comment on spin-rotation-gravity coupling. Physical Review Letters, 68(25), 3809-3810. https://doi.org/10.1103/physrevlett.68.3809
  27. [30]Ahmadov, A. I., Abasova, K. H., amp; Orucova, M. S. (2021). Bound State Solution schrodinger equation for extended Cornell potential at finite temperature. Advances in High Energy Physics, 2021, 1–13. https://doi.org/10.1155/2021/1861946
  28. [31]C. Y. Wong. Introduction to High-Energy Heavy-Ion Collisions, World Scientific Publisher, 1994. https://doi.org/10.1142/1128
  29. [32]F. Becattini, V. Chandra, L. Del Zanna and E. Grossi, Annals Phys. 338 (2013), 32-49 doi:10.1016/j.aop.2013.07.004 [arXiv:1303.3431 [nucl-th]].DOI
  30. [33]G. Torrieri, Phys. Rev. C 98, no.1, 014901 (2018) doi:10.1103/PhysRevC.98.014901 [arXiv:1802.09011 [nucl-th]].DOI
  31. [34]P. Faccioli, V. Knünz, C. Lourenco, J. Seixas and H. K. WöhrI, Phys. Lett. B 736 (2014), 98-109 doi:10.1016/j.physletb.2014.07.006 [arXiv:1403.3970 [hep-ph]].DOI
  32. [35]V. Cheung and R. Vogt, [arXiv:2203.10154 [hep-ph]].DOI
  33. [36]J. Adams, A. Ewigleben, S. Garrett, W. He, T. C. Huang, P. M. Jacobs, X. Ju, M. A. Lisa, M. Lomnitz and R. Pak, et al. Nucl. Instrum. Meth. A 968 (2020), 163970 doi:10.1016/j.nima.2020.163970 [arXiv:1912.05243 [physics.ins-det]]DOI
  34. [37]G. Giacalone, L. Yan, J. Noronha-Hostler and J. Y. Ollitrault, Phys. Rev. C 94 (2016) no.1, 014906 doi:10.1103/PhysRevC.94.014906 [arXiv:1605.08303 [nucl-th]].DOI
  35. [38]W. Florkowski, A. Kumar, R. Ryblewski and A. Mazeliauskas, Phys. Rev. C 100 (2019) no.5, 054907 doi:10.1103/PhysRevC.100.054907 [arXiv:1904.00002 [nucl-th]].DOI
  36. [39]K. J. Gonçalves and G. Torrieri, Baryon polarization and Spin alignment of vector mesons in a thermal model with spin-vorticity non-equilibrium, to be submitted

Paper details

vlnc:W_s7z6xjen
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