betaFeedback

WorksPhysics and Astronomy

Hybrid framework of Fermi-Dirac spin hydrodynamics

Abstract

We outline the hybrid framework of spin hydrodynamics, combining classical kinetic theory with the Israel-Stewart method of introducing dissipation. We obtain the local equilibrium expressions for the baryon current, the energy-momentum tensor, and the spin tensor of particles with spin 1/2 following the Fermi-Dirac statistics and compare them with the previously derived versions where the Boltzmann approximation was used. The expressions in the two cases have the same form, but the coefficients are governed by different functions. The relative differences between the tensor coefficients in the Fermi-Dirac and Boltzmann cases are found to grow exponentially with the baryon chemical potential. In the proposed formalism, nonequilibrium processes are studied including mathematically possible dissipative corrections. Standard conservation laws are applied, and the condition of positive entropy production allows for transfer between the spin and orbital parts of angular momentum.

Introduction

Experimental results indicating a nonzero spin polarization of particles produced in heavy-ion collisions [1–4] have generated a growing interest in observables related to spin degrees of freedom [5], motivating at the same time attempts at expanding the theoretical framework of relativistic hydrodynamics [6, 7] to include a description of spin, thus giving rise to the field of spin hydrodynamics. This challenge of developing a new theoretical formalism has been pursued along several paths based on different assumptions: (a) The final particle spin polarization is determined on the basis of gradients of hydrodynamic fields on the freezeout hypersurface. The vorticity ϖμν=−12(∂μβν−∂νβμ)\varpi_{\mu\nu} = -\frac{1}{2}(\partial_{\mu}\beta_{\nu} - \partial_{\nu}\beta_{\mu}) and the thermal shear ξμν=−12(∂μβν+∂νβμ)\xi_{\mu\nu} = -\frac{1}{2}(\partial_{\mu}\beta_{\nu} + \partial_{\nu}\beta_{\mu}), where βμ=uμT\beta^{\mu} = \frac{u^{\mu}}{T}, are considered to be of particular importance in this case, see Refs. [8–12]. (b) The spin hydrodynamics equations are derived from quantum or classical kinetic theory of particles with spin 1/2 [16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28] [13, 14, 15]. (c) Following the method of Israel and Stewart [29], the laws of conservation and entropy production are applied to general mathematically allowed forms of the energy-momentum and spin tensors [30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45].

(d) Lagrangian effective field theory techniques are used in the context of a spin-polarizable medium [46, 47, 48].

We believe that it is useful and important to reexamine the assumptions and establish relations between the formalisms. Any clarification and synthesis thus achieved may facilitate further progress and, perhaps, eventually lead to removing the present discrepancies. Hence the recent proposal of a hybrid framework that combines the use of kinetic theory for the perfect fluid description and the Israel-Stewart method for the inclusion of dissipation [49, 50, 51]. Among the attractive features of this approach are the simple definition of local thermodynamic equilibrium of particles with spin (conservation of the spin part of the total angular momentum), consistent and independent power counting in the expansion in the spin polarization tensor ωμν\omega^{\mu\nu} and in the gradients (the latter appear in dissipative terms, while the second order expansion in the former is necessary already on the perfect-fluid level to obtain nontrivial thermodynamic relations with spin), as well as the addition of dissipative effects without resorting to the complicated formalism of nonlocal collisions [19, 20, 21, 22, 23].

In this contribution, we summarize the hydrodynamic tensors appearing in the hybrid framework, expand the analysis of the more realistic case of the Fermi-Dirac statistics and compare the results with the previously derived Boltzmann approximation.

Results

Tensors of perfect spin hydrodynamics

We follow the prescription in Ref. [50] to derive the tensors of perfect spin hydrodynamics via the kinetic theory. We focus in particular on the case of the Fermi-Dirac statistics. In the kinetic theory with classical treatment of spin, one introduces the internal angular momentum [52, 53]

sαβ=1mϵαβγδpγsδ,(1)s^{\alpha\beta} = \frac{1}{m}\epsilon^{\alpha\beta\gamma\delta}p_{\gamma}s_{\delta}, \tag*{(1)}

where the spin 4-vector sμs^\mu is orthogonal to the 4-momentum s⋅p=0s \cdot p = 0, sα=12mϵαβγδpβsγδs^\alpha= \frac{1}{2m}\epsilon^{\alpha\beta\gamma\delta}p_\beta s_{\gamma\delta}. In the particle rest frame, pμ=(m,0,0,0)p^\mu= (m,0,0,0), sα=(0,s∗)s^\alpha= (0,\mathbf{s}_*), ∣s∗∣=s|\mathbf{s}_*| = \mathfrak{s}, and s2=1/2(1+1/2)=3/4\mathfrak{s}^2 = 1/2(1+1/2) = 3/4. The local equilibrium distribution functions for particles (+) and for antiparticles (−) have the Fermi-Dirac form

Neqμ=(n0+n2ω+n2k)uμ+nttμN^\mu_{\mathrm{eq}}=(n_0+n_2^\omega+n_2^k)u^\mu+n_t t^\mu
CoefficientFermi-DiracBoltzmann
n0n_0m3π2J21+\frac{m^3}{\pi^2}J_{21}^{+}2sinh⁡ξπ2z2T3K2(z)\frac{2\sinh\xi}{\pi^2}z^2T^3K_2(z)
n2ωn_2^\omega−ω2ξ2m318π2(J˙21++2J˙23+)-\omega^2\frac{\xi^2m^3}{18\pi^2}\left(\dot{J}_{21}^{+}+2\dot{J}_{23}^{+}\right)−ω2ξ2sinh⁡ξ3π2zT3[zK2(z)+2K3(z)]-\omega^2\frac{\xi^2\sinh\xi}{3\pi^2}zT^3\left[zK_2(z)+2K_3(z)\right]
n2kn_2^k−k2ξ2m39π2J˙41+-k^2\frac{\xi^2m^3}{9\pi^2}\dot{J}_{41}^{+}−k22ξ2sinh⁡ξ3π2zT3K3(z)-k^2\frac{2\xi^2\sinh\xi}{3\pi^2}zT^3K_3(z)
ntn_t−ξ2m39π2J˙41+-\frac{\xi^2m^3}{9\pi^2}\dot{J}_{41}^{+}−2ξ2sinh⁡ξ3π2zT3K3(z)-\frac{2\xi^2\sinh\xi}{3\pi^2}zT^3K_3(z)

Table 1. The baryon current

feq±(x,p,s)=[exp⁡(∓ξ(x)+p⋅β(x)−12ω(x):s)+1]−1=1ey±+1.(2)f_{\mathrm{eq}}^{\pm}(x,p,s)=\left[\exp\left(\mp\xi(x)+p\cdot\beta(x)-\frac{1}{2}\omega(x):s\right)+1\right]^{-1}=\frac{1}{e^{y^{\pm}}+1}. \tag*{(2)}

with

y±=∓ξ(x)+p⋅β(x)−12ω(x):s.(3)y^{\pm}=\mp\xi(x)+p\cdot\beta(x)-\frac{1}{2}\omega(x):s. \tag*{(3)}

The colon denotes contraction over both indices, ω:s≡ωμνsμν\omega:s\equiv\omega_{\mu\nu}s^{\mu\nu}. The spin polarization tensor is parametrized with the help of 4-vectors kμk^\mu and tμt^\mu,

ωαβ=kαuβ−kβuα+tαβ,(4)\omega_{\alpha\beta}=k_\alpha u_\beta-k_\beta u_\alpha+t_{\alpha\beta}, \tag*{(4)}

with k⋅u=ω⋅u=0k\cdot u=\omega\cdot u=0, tαβ=ϵαβγδuγωδt_{\alpha\beta}=\epsilon_{\alpha\beta\gamma\delta}u^\gamma\omega^\delta, tμ=tμνkν=ϵμναβkνuαωβt^\mu=t^{\mu\nu}k_\nu=\epsilon^{\mu\nu\alpha\beta}k_\nu u_\alpha\omega_\beta.

In local equilibrium, the baryon current, the energy-momentum tensor, and the spin tensor can be written as

Neqμ=∫dP dS pμ[feq+(x,p,s)−feq−(x,p,s)],(5)N^\mu_{\mathrm{eq}}=\int dP\,dS\,p^\mu\left[f_{\mathrm{eq}}^{+}(x,p,s)-f_{\mathrm{eq}}^{-}(x,p,s)\right], \tag*{(5)}
Teqμν=∫dP dS pμpν[feq+(x,p,s)+feq−(x,p,s)],(6)T^{\mu\nu}_{\mathrm{eq}}=\int dP\,dS\,p^\mu p^\nu\left[f_{\mathrm{eq}}^{+}(x,p,s)+f_{\mathrm{eq}}^{-}(x,p,s)\right], \tag*{(6)}
Seqλ,μν=∫dP dS pλsμν[feq+(x,p,s)+feq−(x,p,s)],(7)S_{\mathrm{eq}}^{\lambda,\mu\nu}=\int dP\,dS\,p^\lambda s^{\mu\nu}\left[f_{\mathrm{eq}}^{+}(x,p,s)+f_{\mathrm{eq}}^{-}(x,p,s)\right], \tag*{(7)}

with the integration measures

dP=d3p(2π)3Ep,dS=mπℏds δ(s⋅s+ℏ2)δ(p⋅s).(8)dP=\frac{d^3p}{(2\pi)^3E_p},\qquad dS=\frac{m}{\pi\hbar}ds\,\delta(s\cdot s+\hbar^2)\delta(p\cdot s). \tag*{(8)}

For small magnitudes of ωμν\omega^{\mu\nu}, the Fermi-Dirac distribution function (2) can be expanded around ys≡−12ω:sy_s \equiv-\frac{1}{2}\omega:s, s=0s=0

feq±(x,p,s)=1ey0±+ys+1=1ey0±+1−ey0±(ey0±+1)2ys+ey0±(ey0±−1)2(ey0±+1)3ys2+⋯ ,(9)f_{\mathrm{eq}}^{\pm}(x,p,s)=\frac{1}{e^{y_0^\pm+y_s}+1}=\frac{1}{e^{y_0^\pm}+1}-\frac{e^{y_0^\pm}}{(e^{y_0^\pm}+1)^2}y_s+\frac{e^{y_0^\pm}(e^{y_0^\pm}-1)}{2(e^{y_0^\pm}+1)^3}y_s^2+\cdots, \tag*{(9)}

where y0±≡∓ξ(x)+p⋅β(x)y_0^\pm\equiv\mp\xi(x)+p\cdot\beta(x). Thus,

feq±=f0±−f0±(1−f0±)ys+12f0±(1−f0±)(1−2f0±)ys2+⋯≡f0±+12f1±ω:s+18f2±(ω:s)2+⋯(10)\begin{aligned} f_{\mathrm{eq}}^\pm=f_0^\pm-f_0^\pm(1-f_0^\pm)y_s+\frac{1}{2}f_0^\pm(1-f_0^\pm)(1-2f_0^\pm)y_s^2+\cdots\\ &\equiv f_0^\pm+\frac{1}{2}f_1^\pm\omega:s+\frac{1}{8}f_2^\pm(\omega:s)^2+\cdots \tag*{(10)} \end{aligned}

Expansion up to the quadratic order in ωμν\omega^{\mu\nu} in the case of NμN^\mu and TμνT^{\mu\nu} (the linear terms vanish), and up to the linear order in ωμν\omega^{\mu\nu} in the case of Sλ,μνS^{\lambda,\mu\nu} yields

Neqμ=∫dP dS pμ[feq+(x,p,s)−feq−(x,p,s)]=2(Z0+μ−Z0−μ)+ℏ2ω:ω6(Z2+μ−Z2−μ)+ℏ23m2(Z2+μαβ−Z2−μαβ)ωγαωβγ+⋯(11)N_{\mathrm{eq}}^\mu=\int dP\,dS\,p^\mu\left[f_{\mathrm{eq}}^+(x,p,s)-f_{\mathrm{eq}}^-(x,p,s)\right]=2(Z_0^{+\mu}-Z_0^{-\mu}) +\hbar^2\frac{\omega:\omega}{6}(Z_2^{+\mu}-Z_2^{-\mu})+\frac{\hbar^2}{3m^2}(Z_2^{+\mu\alpha\beta}-Z_2^{-\mu\alpha\beta})\omega^\gamma{}_\alpha\omega_{\beta\gamma}+\cdots \tag*{(11)}
Teqμν=∫dP dS pμpν[feq+(x,p,s)+feq−(x,p,s)]=2(Z0+μν+Z0−μν)+ℏ2ω:ω6(Z2+μν+Z2−μν)+ℏ23m2(Z2+μναβ+Z2−μναβ)ωγαωβγ+⋯(12)T_{\mathrm{eq}}^{\mu\nu}=\int dP\,dS\,p^\mu p^\nu\left[f_{\mathrm{eq}}^+(x,p,s)+f_{\mathrm{eq}}^-(x,p,s)\right]=2(Z_0^{+\mu\nu}+Z_0^{-\mu\nu}) +\hbar^2\frac{\omega:\omega}{6}(Z_2^{+\mu\nu}+Z_2^{-\mu\nu})+\frac{\hbar^2}{3m^2}(Z_2^{+\mu\nu\alpha\beta}+Z_2^{-\mu\nu\alpha\beta})\omega^\gamma{}_\alpha\omega_{\beta\gamma}+\cdots \tag*{(12)}
Seqλ,μν=∫dP dS pλsμν[feq+(x,p,s)+feq−(x,p,s)]=2ξ23ωμν(Z1+λ+Z1−λ)+2ξ23m2[(Z1+λαμ+Z1−λαμ)ω αν−(Z1+λαν+Z1−λαν)ω αμ]+⋯(13)S_{\mathrm{eq}}^{\lambda,\mu\nu}=\int\mathrm{d}P\,\mathrm{d}S\,p^\lambda s^{\mu\nu}\left[f_{\mathrm{eq}}^+(x,p,s)+f_{\mathrm{eq}}^-(x,p,s)\right]=\frac{2\xi^2}{3}\omega^{\mu\nu}\left(Z_1^{+\lambda}+Z_1^{-\lambda}\right) +\frac{2\xi^2}{3m^2}\left[\left(Z_1^{+\lambda\alpha\mu}+Z_1^{-\lambda\alpha\mu}\right)\omega^\nu_{\ \alpha}-\left(Z_1^{+\lambda\alpha\nu}+Z_1^{-\lambda\alpha\nu}\right)\omega^\mu_{\ \alpha}\right]+\cdots \tag*{(13)}

with the tensors ZnZ_n defined as

Zn±αβ…≡∫dP pαpβ⋯fn±.(14)Z_n^{\pm\alpha\beta\ldots}\equiv\int\mathrm{d}P\,p^\alpha p^\beta\cdots f_n^\pm. \tag*{(14)}

In the Boltzmann approximation, the local distribution function is simpler, feq±=exp⁡(y±)f_{\mathrm{eq}}^\pm=\exp(y^\pm) (compare Eqs. (2) and (3)), and the tensors ZnZ_n simplify to the form

Zn±αβ…→e±ξZαβ…≡e±ξ∫dPpαpβ⋯e−β⋅p,n=0,1,2,…(15)Z_n^{\pm\alpha\beta\ldots}\to e^{\pm\xi}Z^{\alpha\beta\ldots}\equiv e^{\pm\xi}\int\mathrm{d}Pp^\alpha p^\beta\cdots e^{-\beta\cdot p},\qquad n=0,1,2,\ldots \tag*{(15)}

The tensors Z0Z_0, Z1Z_1 and Z2Z_2 with up to four indices were expressed in terms of the function

Jmn(ξ,z)=∫0∞sinh⁡mycosh⁡nyexp⁡(−ξ+zcosh⁡y)+1 dy,m,n=0,1,2,…(16)J_{mn}(\xi,z)=\int_0^\infty\frac{\sinh^m y\cosh^n y}{\exp(-\xi+z\cosh y)+1}\,\mathrm{d}y,\qquad m,n=0,1,2,\ldots \tag*{(16)}

and its derivatives in Appendix A of Ref. [50], although subsequently only the Boltzmann approximation was considered. Here we address the Fermi-Dirac case. We denote the first and the second derivative of the function JmnJ_{mn} with respect to ξ\xi with J˙mn\dot{J}_{mn} and J¨mn\ddot{J}_{mn}, respectively

J˙mn(ξ,z)≡∂∂ξJmn(ξ,z)=∫0∞sinh⁡mycosh⁡ny2+2cosh⁡(ξ−zcosh⁡y) dy,\dot{J}_{mn}(\xi,z)\equiv\frac{\partial}{\partial\xi}J_{mn}(\xi,z)=\int_0^\infty\frac{\sinh^m y\cosh^n y}{2+2\cosh(\xi-z\cosh y)}\,\mathrm{d}y,
J¨mn(ξ,z)≡∂2∂ξ2Jmn(ξ,z)=∫0∞sinh⁡mycosh⁡nysinh⁡(ξ−zcosh⁡y)2(1+cosh⁡(ξ−zcosh⁡y))2 dy.(17)\ddot{J}_{mn}(\xi,z)\equiv\frac{\partial^2}{\partial\xi^2}J_{mn}(\xi,z)=\int_0^\infty\frac{\sinh^m y\cosh^n y\sinh(\xi-z\cosh y)}{2(1+\cosh(\xi-z\cosh y))^2}\,\mathrm{d}y. \tag*{(17)}

For brevity, we introduce the following notation using the plus and minus superscript

Jmn+≡Jmn(ξ,z),Jmn−≡Jmn(−ξ,z),J^{+}_{mn} \equiv J_{mn}(\xi,z),\qquad J^{-}_{mn} \equiv J_{mn}(-\xi,z),
Jmn±≡Jmn++Jmn−,Jmn∓≡Jmn+−Jmn−(18)J^{\pm}_{mn} \equiv J^{+}_{mn}+J^{-}_{mn},\qquad J^{\mp}_{mn} \equiv J^{+}_{mn}-J^{-}_{mn} \tag*{(18)}

and the same for J˙\dot{J} and J¨\ddot{J}.

We gather the expressions derived in Appendix A of Ref. [50], insert them into (11)–(13), and perform contractions and further simplifications, also noting the usefulness of hyperbolic identities in expressing the functions JmnJ_{mn}, J˙mn\dot{J}_{mn} and J¨mn\ddot{J}_{mn} by functions with different indices, as needed to convert the expressions into their most compact forms. Finally, we can write the three tensors as

Neqμ=(n0+n2ω+n2k)uμ+nttμ,(19)N_{\mathrm{eq}}^{\mu}=(n_{0}+n_{2}^{\omega}+n_{2}^{k})u^{\mu}+n_{t}t^{\mu}, \tag*{(19)}
Teqμν=(ϵ0+ϵ2ω+ϵ2k)uμuν−(P0+P2ω+P2k)Δμν+Pt(tμuν+tνuμ)+Pkω(kμkν+ωμων),(20)T_{\mathrm{eq}}^{\mu\nu}=(\epsilon_{0}+\epsilon_{2}^{\omega}+\epsilon_{2}^{k})u^{\mu}u^{\nu}-(P_{0}+P_{2}^{\omega}+P_{2}^{k})\Delta^{\mu\nu} +P_{t}(t^{\mu}u^{\nu}+t^{\nu}u^{\mu})+P_{k\omega}(k^{\mu}k^{\nu}+\omega^{\mu}\omega^{\nu}), \tag*{(20)}
Seqλ,μν=uλ[A(kμuν−kνuμ)+A1tμν]+A2(tλμuν−tλνuμ+Δλμkν−Δλνkμ),(21)S_{\mathrm{eq}}^{\lambda,\mu\nu}=u^{\lambda}\left[A(k^{\mu}u^{\nu}-k^{\nu}u^{\mu})+A_{1}t^{\mu\nu}\right]+\frac{A}{2}\left(t^{\lambda\mu}u^{\nu}-t^{\lambda\nu}u^{\mu}+\Delta^{\lambda\mu}k^{\nu}-\Delta^{\lambda\nu}k^{\mu}\right), \tag*{(21)}

with Δμν≡gμν−uμuν\Delta^{\mu\nu}\equiv g^{\mu\nu}-u^{\mu}u^{\nu}. Their form is the same in the Fermi-Dirac and the Boltzmann case, but the functional dependence of the coefficients on mm, TT and μ\mu differs, as the simpler integrals appearing in Eq. (15) can be expressed in terms of Bessel functions of the second kind Kn(z)K_{n}(z). We note that compared with the standard perfect-fluid expressions,

Neqμ=nuμ,(22)N_{\mathrm{eq}}^{\mu}=nu^{\mu}, \tag*{(22)}
Teqμν=ϵuμuν−PΔμν,(23)T_{\mathrm{eq}}^{\mu\nu}=\epsilon u^{\mu}u^{\nu}-P\Delta^{\mu\nu}, \tag*{(23)}

and the “phenomenological” form of the spin tensor [54],

Seqλ,μν=uλSμν,(24)S_{\mathrm{eq}}^{\lambda,\mu\nu}=u^{\lambda}S^{\mu\nu}, \tag*{(24)}

there are additional terms. Although seemingly their mathematical structure is typical of dissipative corrections, in this framework they appear already in the description of the perfect fluid; the entropy current, defined here as

Seqμ=Tμαβα−12ωαβSeqμ,αβ−ξNeqμ+Nμ,Nμ=coth⁡ξ Nμ,(25)S_{\mathrm{eq}}^{\mu}=T^{\mu\alpha}\beta_{\alpha}-\frac{1}{2}\omega_{\alpha\beta}S_{\mathrm{eq}}^{\mu,\alpha\beta}-\xi N_{\mathrm{eq}}^{\mu}+\mathcal{N}^{\mu},\qquad\mathcal{N}^{\mu}=\coth\xi\,N^{\mu}, \tag*{(25)}

remains conserved. Our complete results for the Fermi-Dirac and Boltzmann tensor coefficients have been collected in Tables I, Table II, and Table III for the baryon current, the energy-momentum tensor, and the spin tensor, respectively.

Teqμν=(ε0+ε2ω+ε2k)uμuν−(P0+P2ω+P2k)Δμν+Pt(tμuν+tνuμ)+Pkω(kμkν+ωμων)T_{\mathrm{eq}}^{\mu\nu} = (\varepsilon_0 + \varepsilon_2^\omega+ \varepsilon_2^k)u^\mu u^\nu- (P_0 + P_2^\omega+ P_2^k)\Delta^{\mu\nu} + P_t(t^\mu u^\nu+ t^\nu u^\mu) + P_{k\omega}(k^\mu k^\nu+ \omega^\mu\omega^\nu)
CoefficientFermi-DiracBoltzmann
ε0\varepsilon_0m4π2J22±\frac{m^4}{\pi^2}J_{22}^{\pm}2cosh⁡ξπ2zT4[zK3(z)−K2(z)]\frac{2\cosh\xi}{\pi^2}zT^4[zK_3(z)-K_2(z)]
ε2ω\varepsilon_2^\omega−ω2ℏ2m418π2(J¨24±+2J¨24±)-\omega^2\frac{\hbar^2m^4}{18\pi^2}(\ddot{J}_{24}^{\pm}+2\ddot{J}_{24}^{\pm})−ω2ℏ2cosh⁡ξ3π2zT4[zK2(z)+(z2+10)K3(z)]-\omega^2\frac{\hbar^2\cosh\xi}{3\pi^2}zT^4[zK_2(z)+(z^2+10)K_3(z)]
ε2k\varepsilon_2^k−k2ℏ2m49π2J¨42±-k^2\frac{\hbar^2m^4}{9\pi^2}\ddot{J}_{42}^{\pm}−k24ℏ2cosh⁡ξ3π2zT4[zK2(z)+5K3(z)]-k^2\frac{4\hbar^2\cosh\xi}{3\pi^2}zT^4[zK_2(z)+5K_3(z)]
P0P_0m43π2J40±\frac{m^4}{3\pi^2}J_{40}^{\pm}2cosh⁡ξπ2z2T4K2(z)\frac{2\cosh\xi}{\pi^2}z^2T^4K_2(z)
P2ωP_2^\omega−ω2ℏ2m490π2(J¨40±+4J¨42±)-\omega^2\frac{\hbar^2m^4}{90\pi^2}(\ddot{J}_{40}^{\pm}+4\ddot{J}_{42}^{\pm})−ω2ℏ2cosh⁡ξ3π2zT4[zK2(z)+4K3(z)]-\omega^2\frac{\hbar^2\cosh\xi}{3\pi^2}zT^4[zK_2(z)+4K_3(z)]
P2kP_2^k−k22ℏ2m445π2J¨60±-k^2\frac{2\hbar^2m^4}{45\pi^2}\ddot{J}_{60}^{\pm}−k24ℏ2cosh⁡ξ3π2zT4K3(z)-k^2\frac{4\hbar^2\cosh\xi}{3\pi^2}zT^4K_3(z)
PkωP_{k\omega}−ℏ2m445π2J¨60±-\frac{\hbar^2m^4}{45\pi^2}\ddot{J}_{60}^{\pm}−2ℏ2cosh⁡ξ3π2zT4K3(z)-\frac{2\hbar^2\cosh\xi}{3\pi^2}zT^4K_3(z)
PtP_t−ℏ2m49π2J¨42±-\frac{\hbar^2m^4}{9\pi^2}\ddot{J}_{42}^{\pm}−2ℏ2cosh⁡ξ3π2zT4[zK2(z)+5K3(z)]-\frac{2\hbar^2\cosh\xi}{3\pi^2}zT^4[zK_2(z)+5K_3(z)]

Table II. The energy-momentum tensor

CoefficientFermi-DiracBoltzmann
A1A_1ξ2m39π2(J˙21±+2J˙23±)\frac{\xi^2m^3}{9\pi^2}(\dot{J}_{21}^{\pm}+2\dot{J}_{23}^{\pm})2ξ2cosh⁡ξ3π2zT3[zK2(z)+2K3(z)]\frac{2\xi^2\cosh\xi}{3\pi^2}zT^3[zK_2(z)+2K_3(z)]
A2A_22ξ2m33π2(2J˙23±−J˙21±)\frac{2\xi^2m^3}{3\pi^2}(2\dot{J}_{23}^{\pm}-\dot{J}_{21}^{\pm})4ξ2cosh⁡ξ3π2z2T3K4(z)\frac{4\xi^2\cosh\xi}{3\pi^2}z^2T^3K_4(z)
A3A_3−2ξ2m39π2J˙41±-\frac{2\xi^2m^3}{9\pi^2}\dot{J}_{41}^{\pm}−4ξ2cosh⁡ξ3π2zT3K3(z)-\frac{4\xi^2\cosh\xi}{3\pi^2}zT^3K_3(z)

Table III. TABLE III: The spin tensor. Seqλ,μν=A1uλωμν+A22uλ(uμkν−uνkμ)+A32tλμν=uλ[A3(kμuν−kνuμ)+A1tμν]+A32(tλμuν−tλνuμ+Δλμkν−Δλνkμ)S_{\mathrm{eq}}^{\lambda,\mu\nu}=A_1u^\lambda\omega^{\mu\nu}+\frac{A_2}{2}u^\lambda(u^\mu k^\nu-u^\nu k^\mu)+\frac{A_3}{2}t^{\lambda\mu\nu}=u^\lambda[A_3(k^\mu u^\nu-k^\nu u^\mu)+A_1t^{\mu\nu}]+\frac{A_3}{2}(t^{\lambda\mu}u^\nu-t^{\lambda\nu}u^\mu+\Delta^{\lambda\mu}k^\nu-\Delta^{\lambda\nu}k^\mu).

Figure 1 shows a comparison between the Fermi-Dirac and Boltzmann cases. As can be readily seen in the logarithmic graphs, for constant particle mass mm and temperature TT, the relative difference increases exponentially with the baryon chemical potential μ\mu for most of its plotted range. In the low μ\mu regime, where the Fermi-Dirac distribution can be satisfactorily approximated by the Boltzmann distribution, the difference between the coefficient values in the two cases is negligible, but in the high μ\mu regime it is significant. The relative differences between the two cases are largest for n2ωn_{2}^{\omega} and ϵ2ω\epsilon_{2}^{\omega}, reaching the order of 1 for ξ=μ/T≈10\xi= \mu/T \approx10 (for the sample values of m=1000m = 1000 MeV and T=100T = 100 MeV chosen).

Figure 1. Relative differences between the Boltzmann and the Fermi-Dirac case of the coefficients of the baryon current NeqμN_{\rm eq}^{\mu} (a)–(b), the energy-momentum tensor TeqμνT_{\rm eq}^{\mu\nu} (c)–(f) and the spin tensor Seqλ,μνS_{\rm eq}^{\lambda,\mu\nu} (g)–(h). Plots for particle mass m=1000m = 1000 MeV, temperature T=100T = 100 MeV, and baryon chemical potential μ\mu ranging from 0 to 1000 MeV.

Tensors of dissipative spin hydrodynamics

Following the Israel-Stewart method, we write down general nonequilibrium expressions as a sum of the equilibrium terms and dissipative corrections

Nμ=Neqμ+δNμ,Tμν=Teqμν+δTμν,Sμ,αβ=Seqμ,αβ+δSμ,αβ.(26)N^\mu= N^\mu_{\mathrm{eq}} + \delta N^\mu,\qquad T^{\mu\nu} = T^{\mu\nu}_{\mathrm{eq}} + \delta T^{\mu\nu},\qquad S^{\mu,\alpha\beta} = S^{\mu,\alpha\beta}_{\mathrm{eq}} + \delta S^{\mu,\alpha\beta}. \tag*{(26)}

It is through dissipative terms that the spin-orbit interaction is introduced: TμνT^{\mu\nu} can now contain nonsymmetric parts, and the spin tensor is no longer seperately conserved

∂μJμ,αβ=0,Jμ,αβ=xαTμβ−xβTμα+Sμ,αβ,(27)\partial_\mu J^{\mu,\alpha\beta} = 0,\qquad J^{\mu,\alpha\beta} = x^\alpha T^{\mu\beta} - x^\beta T^{\mu\alpha} + S^{\mu,\alpha\beta}, \tag*{(27)}
∂μSμ,αβ=Tβα−Tαβ,∂μNμ=0,∂μTμν=0.(28)\partial_\mu S^{\mu,\alpha\beta} = T^{\beta\alpha} - T^{\alpha\beta},\qquad\partial_\mu N^\mu= 0,\qquad\partial_\mu T^{\mu\nu} = 0. \tag*{(28)}

From Eqs. (25), (26), (28), we get

∂μSμ=δNμ∂μξ+δTsμν∂μβν+δTaμν(∂μβν−ωνμ)−12δSμ,αβ∂μωαβ;(29)\partial_\mu S^\mu= \delta N^\mu\partial_\mu\xi+ \delta T^{\mu\nu}_{s}\partial_\mu\beta_\nu+ \delta T^{\mu\nu}_{a}\left(\partial_\mu\beta_\nu- \omega_{\nu\mu}\right) - \frac{1}{2}\delta S^{\mu,\alpha\beta}\partial_\mu\omega_{\alpha\beta}; \tag*{(29)}

compare, e.g., [30, 55]. To find the form of the deviations δNμ\delta N^\mu, δTμν\delta T^{\mu\nu}, δSμ,αβ\delta S^{\mu,\alpha\beta}, we use a decomposition of general tensors of the given symmetry via projections along uμu^\mu and separation into symmetric and antisymmetric parts, obtaining, through the Landau matching conditions (see [50] for details),

δNμ=Vμ,\delta N^\mu= V^\mu,
δTsμν=−ΠΔμν+Wμuν+Wνuμ+πμν,\delta T^{\mu\nu}_{s} = -\Pi\Delta^{\mu\nu} + W^\mu u^\nu+ W^\nu u^\mu+ \pi^{\mu\nu},
δTaμν=dμuν−dνuμ+eμν,\delta T^{\mu\nu}_{a} = d^\mu u^\nu- d^\nu u^\mu+ e^{\mu\nu},
δSλ,μν=Σλμuν−Σλνuμ+ϕλμν,(30)\delta S^{\lambda,\mu\nu} = \Sigma^{\lambda\mu}u^\nu- \Sigma^{\lambda\nu}u^\mu+ \phi^{\lambda\mu\nu}, \tag*{(30)}

with

Vμ=bμ−nttμ,Π=e−(P0+P2ω+P2k)+13Pkω(k2+ω2),Wμ=dsμ−Pttμ,πμν=es⟨μν⟩−Pkω(k⟨μkν⟩+ω⟨μων⟩),Σλμ=iλμ−A2tλμ,ϕλμν=jλμν−A2(Δλμkν−Δλνkμ).(31)\begin{aligned} V^\mu&= b^\mu-n_t t^\mu,\qquad\Pi=e-\left(P_0+P_2^\omega+P_2^k\right)+\frac{1}{3}P_{k\omega}\left(k^2+\omega^2\right),\\ W^\mu&= d_s^\mu-P_t t^\mu,\qquad\pi^{\mu\nu}=e_s^{\langle\mu\nu\rangle}-P_{k\omega}\left(k^{\langle\mu}k^{\nu\rangle}+\omega^{\langle\mu}\omega^{\nu\rangle}\right),\\ \Sigma^{\lambda\mu} &= i^{\lambda\mu}-\frac{A}{2}t^{\lambda\mu},\qquad\phi^{\lambda\mu\nu}=j^{\lambda\mu\nu}-\frac{A}{2}\left(\Delta^{\lambda\mu}k^\nu-\Delta^{\lambda\nu}k^\mu\right). \tag*{(31)} \end{aligned}

Angular brackets around a pair of indices denote the orthogonal, symmetric, and traceless part of the tensor, defined via a contraction with the projector Δαβμν\Delta^{\mu\nu}_{\alpha\beta}, A⟨αβ⟩≡ΔμναβAμν≡12(ΔμαΔνβ+ΔναΔμβ−23ΔμνΔαβ)AαβA^{\langle\alpha\beta\rangle}\equiv\Delta^{\alpha\beta}_{\mu\nu}A^{\mu\nu}\equiv\frac{1}{2}\left(\Delta^\alpha_\mu\Delta^\beta_\nu+\Delta^\alpha_\nu\Delta^\beta_\mu-\frac{2}{3}\Delta^{\mu\nu}\Delta_{\alpha\beta}\right)A^{\alpha\beta}. Squared brackets denote the antisymmetric part, A[μν]≡12(Aμν−Aνμ)A^{[\mu\nu]}\equiv\frac{1}{2}\left(A^{\mu\nu}-A^{\nu\mu}\right). Angular brackets around a single index denote an orthogonal projection Tαβ⟨γ⟩δ…≡ΔργTαβρδ…T^{\alpha\beta\langle\gamma\rangle\delta\ldots}\equiv\Delta^\gamma_\rho T^{\alpha\beta\rho\delta\ldots}. Differential operators take precedence over the orthogonal projection. The nabla symbol denotes the transverse gradient ∇μ≡Δμν∂ν\nabla^\mu\equiv\Delta^{\mu\nu}\partial_\nu.

Equations (30) have a form that was analyzed in Ref. [37], and the coefficients bμb^\mu, dsμd_s^\mu, daμd_a^\mu, ee, es⟨μν⟩e_s^{\langle\mu\nu\rangle}, eaμνe_a^{\mu\nu}, iλμi^{\lambda\mu}, jλμνj^{\lambda\mu\nu} can be expressed in terms of gradients of hydrodynamic variables multiplied by kinetic coefficients, including the shear and bulk viscosity η,ζ\eta,\zeta, the thermal conductivity κ\kappa, the coefficients λa\lambda_a and γ\gamma from Ref. [30], and the coefficients χ1,χ2,χ3,χ4\chi_1,\chi_2,\chi_3,\chi_4 introduced in Ref. [37]. To the first order in gradients, we find

bμ=λ∇μξ+nttμ,dsμ=−κ(Duμ−β∇μT)+Pttμ,daμ=λaβ−1(βDuμ+β2∇μT−2kμ),e=Pˉ−ζθ−13Pkω(k2+ω2),es⟨μν⟩=2ησμν+Pkω(k⟨μkν⟩+ω⟨μων⟩),eaμν=γβ∇[μuν],iλμ=−χ1Δλμuβ∇αωαβ−χ2uν∇⟨λωμ⟩ν−χ3uνΔρ[μ∇λ]ωρν+A2tλμ,jλμν=χ42∇⟨μωλν⟩+A2(Δλμkν−Δλνkμ).(32)\begin{aligned} b^\mu&= \lambda\nabla^\mu\xi+n_t t^\mu,\qquad d_s^\mu=-\kappa\left(Du^\mu-\beta\nabla^\mu T\right)+P_t t^\mu,\\ d_a^\mu&= \lambda_a\beta^{-1}\left(\beta Du^\mu+\beta^2\nabla^\mu T-2k^\mu\right),\qquad e=\bar{P}-\zeta\theta-\frac{1}{3}P_{k\omega}\left(k^2+\omega^2\right),\\ e_s^{\langle\mu\nu\rangle} &= 2\eta\sigma^{\mu\nu}+P_{k\omega}\left(k^{\langle\mu}k^{\nu\rangle}+\omega^{\langle\mu}\omega^{\nu\rangle}\right),\qquad e_a^{\mu\nu}=\gamma\beta\nabla^{[\mu}u^{\nu]},\\ i^{\lambda\mu} &= -\chi_1\Delta^{\lambda\mu}u^\beta\nabla^\alpha\omega_{\alpha\beta}-\chi_2u_\nu\nabla^{\langle\lambda}\omega^{\mu\rangle\nu}-\chi_3u_\nu\Delta^{[\mu}_{\rho}\nabla^{\lambda]}\omega^{\rho\nu}+\frac{A}{2}t^{\lambda\mu},\\ j^{\lambda\mu\nu} &= \frac{\chi_4}{2}\nabla^{\langle\mu}\omega^{\lambda\nu\rangle}+\frac{A}{2}\left(\Delta^{\lambda\mu}k^\nu-\Delta^{\lambda\nu}k^\mu\right). \tag*{(32)} \end{aligned}

Importantly, the spin polarization tensor ωμν\omega_{\mu\nu}, which in natural units is a dimensionless quantity, is not considered to be a gradient term. Already, the perfect-fluid case is based on an expansion in ωμν\omega_{\mu\nu} to the second order. Gradient terms are only introduced when considering dissipation, and the order in gradients is counted separately to the order in ωμν\omega_{\mu\nu}. Therefore, in the expansion above, which is linear in gradients, terms such as ∇αωαβ\nabla^\alpha\omega_{\alpha\beta} survive and are not neglected as higher-order contributions.

Discussion

The hybrid framework of spin hydrodynamics combines the perfect-fluid results of kinetic theory for particles with spin 1/21/2 with the Israel-Stewart approach for including nonequilibrium processes. It involves a two-fold expansion: in the spin polarization tensor ωμν\omega_{\mu\nu} (already in local equilibrium) and in gradients of hydrodynamic variables (when considering close-to-equilibrium dynamics). If the spin polarization is nonvanishing, the perfect-fluid description contains seemingly dissipative, transverse terms. However, genuine dissipative terms appear only at the level of the dissipative fluid. They are determined by the condition of positive entropy production.

This contribution focuses on summarizing the derivation of the tensors appearing in the hybrid framework, with the emphasis on Fermi-Dirac statistics. Thus, it complements the outline given in Ref. [51], which discusses in more detail the motivation behind the framework, as well as the modification of the usual thermodynamic relations of relativistic hydrodynamics that makes them consistent with the fact that the spin tensor (21) derived from kinetic theory, in contrast with the “phenomenological” spin tensor (24), contains parts orthogonal to the flow vector uμu^\mu.

We have shown that the difference between the tensor coefficients derived for the Fermi-Dirac statistics and their Boltzmann approximation is significant in the regime of high baryon chemical potential.

The coefficients derived herein can be employed in lieu of the Boltzmann approximation in future computer simulations, thereby enhancing their realism. Although the functions JmnJ_{mn}, J˙mn\dot{J}_{mn} and J¨mn\ddot{J}_{mn}, unlike Bessel functions, are not a part of standard mathematical packages, their suitable tabulation and interpolation for use in practical codes presents very little technical difficulty.

The first numerical test of the hybrid framework was performed in a simple boost-invariant perfect-fluid case in Ref. [56]. That analysis can be extended to include the Fermi-Dirac statistics and dissipative terms. The framework should also be straightforward to implement in numerical simulations of spin hydrodynamics in more realistic expansion models, similar to those presented in Refs. [57, 58].

Acknowledgments

I thank my collaborators Wojciech Florkowski and Mykhailo Hontarenko for clarifying discussions.

References

  1. [16]S. Bhadury, W. Florkowski, A. Jaiswal, A. Kumar, and R. Ryblewski, Phys. Rev. D 103, 014030 (2021), arXiv:2008.10976 [nucl-th].DOI
  2. [17]S. Bhadury, W. Florkowski, A. Jaiswal, A. Kumar, and R. Ryblewski, Phys. Rev. Lett. 129, 192301 (2022), arXiv:2204.01357 [nucl-th].DOI
  3. [18]R. Singh, M. Shokri, and S. M. A. T. Mehr, Nucl. Phys. A 1035, 122656 (2023), arXiv:2202.11504 [hep-ph].DOI
  4. [19]N. Weickgenannt, X.-L. Sheng, E. Speranza, Q. Wang, and D. H. Rischke, Phys. Rev. D 100, 056018 (2019), arXiv:1902.06513 [hep-ph].DOI
  5. [20]N. Weickgenannt, E. Speranza, X.-l. Sheng, Q. Wang, and D. H. Rischke, Phys. Rev. D 104, 016022 (2021), arXiv:2103.04896 [nucl-th].DOI
  6. [21]N. Weickgenannt, E. Speranza, X.-l. Sheng, Q. Wang, and D. H. Rischke, Phys. Rev. Lett. 127, 052301 (2021), arXiv:2005.01506 [hep-ph].DOI
  7. [22]D. Wagner, N. Weickgenannt, and D. H. Rischke, Phys. Rev. D 106, 116021 (2022), arXiv:2210.06187 [nucl-th].DOI
  8. [23]N. Weickgenannt, D. Wagner, E. Speranza, and D. H. Rischke, Phys. Rev. D 106, 096014 (2022), arXiv:2203.04766 [nucl-th].DOI
  9. [24]N. Weickgenannt and J.-P. Blaizot, Phys. Rev. D 109, 056012 (2024), arXiv:2311.15817 [hep-ph].arxiv.org/abs/2311.15817
  10. [25]D. Wagner, M. Shokri, and D. H. Rischke, Phys. Rev. Res. 6, 043103 (2024), arXiv:2405.00533 [nucl-th].DOI
  11. [26]J. Hu, Phys. Rev. D 105, 076009 (2022), arXiv:2111.03571 [hep-ph].DOI
  12. [27]S. Li, M. A. Stephanov, and H.-U. Yee, Phys. Rev. Lett. 127, 082302 (2021), arXiv:2011.12318 [hep-th].DOI
  13. [28]S. Shi, C. Gale, and S. Jeon, Phys. Rev. C 103, 044906 (2021), arXiv:2008.08618 [nucl-th].DOI
  14. [29]W. Israel and J. M. Stewart, Annals Phys. 118, 341 (1979).
  15. [30]K. Hattori, M. Hongo, X.-G. Huang, M. Matsuo, and H. Taya, Phys. Lett. B 795, 100 (2019), arXiv:1901.06615 [hep-th].DOI
  16. [31]K. Fukushima and S. Pu, Phys. Lett. B 817, 136346 (2021), arXiv:2010.01608 [hep-th].DOI
  17. [32]A. Daher, A. Das, W. Florkowski, and R. Ryblewski, Phys. Rev. C 108, 024902 (2023), arXiv:2202.12609 [nucl-th].DOI
  18. [33]A. Daher, A. Das, and R. Ryblewski, Phys. Rev. D 107, 054043 (2023), arXiv:2209.10460 [nucl-th].DOI
  19. [34]G. Sarwar, M. Hasanuzzaman, J. R. Bhatt, H. Mishra, and J.-e. Alam, Phys. Rev. D 107, 054031 (2023), arXiv:2209.08652 [nucl-th].DOI
  20. [35]D.-L. Wang, S. Fang, and S. Pu, Phys. Rev. D 104, 114043 (2021), arXiv:2107.11726 [nucl-th].DOI
  21. [36]R. Biswas, A. Daher, A. Das, W. Florkowski, and R. Ryblewski, Phys. Rev. D 107, 094022 (2023), arXiv:2211.02934 [nucl-th].DOI
  22. [37]R. Biswas, A. Daher, A. Das, W. Florkowski, and R. Ryblewski, Phys. Rev. D 108, 014024 (2023), arXiv:2304.01009 [nucl-th].DOI
  23. [38]X.-Q. Xie, D.-L. Wang, C. Yang, and S. Pu, Phys. Rev. D 108, 094031 (2023), arXiv:2306.13880 [hep-ph].arxiv.org/abs/2306.13880
  24. [39]A. Daher, W. Florkowski, and R. Ryblewski, Phys. Rev. D 110, 034029 (2024), arXiv:2401.07608 [hep-ph].arxiv.org/abs/2401.07608
  25. [40]X. Ren, C. Yang, D.-L. Wang, and S. Pu, Phys. Rev. D 110, 034010 (2024), arXiv:2405.03105 [nucl-th].arxiv.org/abs/2405.03105
  26. [41]A. Daher, W. Florkowski, R. Ryblewski, and F. Taghinavaz, Phys. Rev. D 109, 114001 (2024), arXiv:2403.04711 [hep-ph].arxiv.org/abs/2403.04711
  27. [42]A. D. Gallegos, U. Gürsoy, and A. Yarom, SciPost Phys. 11, 041 (2021), arXiv:2101.04759 [hep-th].DOI
  28. [43]M. Hongo, X.-G. Huang, M. Kaminski, M. Stephanov, and H.-U. Yee, JHEP 11, 150, arXiv:2107.14231 [hep-th].DOI
  29. [44]A. Kumar, D.-L. Yang, and P. Gubler, Phys. Rev. D 109, 054038 (2024), arXiv:2312.16900 [nucl-th].DOI
  30. [45]D. She, A. Huang, D. Hou, and J. Liao, Sci. Bull. 67, 2265 (2022), arXiv:2105.04060 [nucl-th].DOI
  31. [46]D. Montenegro, L. Tinti, and G. Torrieri, Phys. Rev. D 96, 056012 (2017), [Addendum: Phys.Rev.D 96, 079901 (2017)], arXiv:1701.08263 [hep-th].DOI
  32. [47]D. Montenegro and G. Torrieri, Phys. Rev. D 100, 056011 (2019), arXiv:1807.02796 [hep-th].DOI
  33. [48]D. Montenegro and G. Torrieri, Phys. Rev. D 102, 036007 (2020), arXiv:2004.10195 [hep-th].DOI
  34. [49]W. Florkowski and M. Hontarenko, Phys. Rev. Lett. 134, 082302 (2025), arXiv:2405.03263 [hep-ph].DOI
  35. [50]Z. Drogosz, W. Florkowski, and M. Hontarenko, Phys. Rev. D 110, 096018 (2024), arXiv:2408.03106 [hep-ph].DOI
  36. [51]W. Florkowski, Journal of Subatomic Particles and Cosmology 3, 100028 (2025), arXiv:2411.19673 [hep-ph].arxiv.org/abs/2411.19673
  37. [52]M. Mathisson, Acta Phys. Polon. 6, 163 (1937).
  38. [53]M. Mathisson, General Relativity and Gravitation 42, 1011 (2010).
  39. [54]J. Weyssenhoff and A. Raabe, Acta Phys. Polon. 9, 7 (1947).
  40. [55]F. Becattini, A. Daher, and X.-L. Sheng, Phys. Lett. B 850, 138533 (2024), arXiv:2309.05789 [nucl-th].arxiv.org/abs/2309.05789
  41. [56]Z. Drogosz, W. Florkowski, N. Łygan, and R. Rybłewski, Phys. Rev. C 111, 024909 (2025), arXiv:2411.06154 [hep-ph].DOI
  42. [57]S. K. Singh, R. Ryblewski, and W. Florkowski, Phys. Rev. C 111, 024907 (2025), arXiv:2411.19673 [hep-ph].
  43. [58]Sapna, S. K. Singh, and D. Wagner, (2025), arXiv:2503.22552 [hep-ph].DOI

Paper details

vlnc:W_9be3qcp8
Version
v1 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC BY 4.0
Usage

views · 8

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