betaFeedback

WorksPhysics and Astronomy

A novel method for calculating Bose-Einstein correlation functions with Coulomb final-state interaction

Abstract

Measurements of Bose-Einstein correlations played a crucial role in the discovery and the subsequent detailed exploration of the Quark-Gluon-Plasma (QGP) created in high-energy collisions of heavy nuclei. Such measurements gave rise to femtoscopy, a flourishing sub-field of high-energy physics, and there are important new directions to explore and discoveries to be made in the near future. One of these important current topics is the precise investigation of the shape of the correlation functions utilizing Lévy-stable sources. In this paper, we present a novel method of calculating the shape of the two-particle correlation functions, including the Coulomb final-state interaction. This method relies on an exact calculation of a large part of the necessary integrals of the Coulomb wave function and can be utilized to calculate the correlation function for any source function with an easily accessible Fourier transform. In this way, it is eminently applicable to Lévy-stable source functions. In this particular case, we find that the new method is more robust and allows the investigation of a wider parameter range than the previously utilized techniques. We present an easily applicable software package that is ready to use in experimental studies.

Introduction

High energy heavy-ion physics aims at recreating the conditions characteristic to the first few microseconds of our Universe and studying the matter that is present under such conditions: the strongly interacting Quark Gluon Plasma (QGP) [1, 2, 3, 4]. Experiments observe the particles produced in heavy ion collisions in order to draw conclusions about the properties of this matter. An important class of observables is that of momentum correlations of identical particles. In the case of bosonic particles (e.g., pions or kaons), these are called Bose-Einstein-correlations because they arise as a result of the quantum statistical properties of the particles. They are also called HBT correlations after R. Hanbury Brown and R. Q. Twiss [5], who first discovered a related effect in astronomy. These correlations are connected to the space-time geometry of the particle emitting source [6]; thus, they provide a vital tool to study the matter produced in these collisions. These types of measurements have been extensively studied for more than 60 years now [7], especially since the era of the Relativistic Heavy Ion Collider (RHIC) started around the early 2000s [8, 9].

The usual way to experimentally study the source function is to assume a source shape (e.g., a Gaussian distribution) characterized by a handful of parameters, calculate the correlation function arising from such a source, then test this assumption on the measured momentum correlation function. If the calculated correlation shape fits the data well (in terms of statistical acceptability), one can extract the source parameters and investigate their dependence on average momentum, centrality, collision energy, etc. One of the most essential steps in such an analysis is thus the calculation of the correlation function for a given source function.

As detailed below, if one neglects the final state interactions of the produced particles, the correlation function is given by the Fourier transform of the source function, a relatively simple calculational step in most cases. However, in the case of charged particles (such as charged pions, the most frequent target of measurements), the correlation function is given by a more complicated integral transform of the source function. In the process of fitting the source parameters to measurements, a frequently used method has been to calculate the Coulomb integral only for a pre-defined set of the parameter values and derive a “Coulomb correction factor” from it. In this way, however, there is an inherent distortion in the fitting procedure since the Coulomb effect is calculated at some arbitrary parameter values instead of the supposedly final, fitted ones. As experimental measurements and phenomenological descriptions become more and more refined, it becomes absolutely necessary that in the fitting procedure, the Coulomb interaction is considered in a self-consistent way. This necessitates the calculation of the mentioned Coulomb transform of the assumed source function for arbitrary source parameters during the fitting procedure.

The simplest assumption for the source function is that of a Gaussian distribution, and in the past decades, many aspects of Gaussian correlation measurements were explored [10, 11, 12]. However, as the experimental resolution and the accumulated data increased throughout the years, an expectation arose to better describe the correlation function’s shape [13]. One possibility is presented by a spherical harmonics expansion [14]; another is to use a functional form for the source that goes beyond the Gaussian approximation. To that end, Lévy-stable distributions were utilized [15, 16, 17]. In this way, a statistically acceptable successful description of the data was achieved. The utilization of more and more general source function assumptions, such as the mentioned Lévy distributions, is thus another pressing motivation for developing precise calculational tools to treat the Coulomb interaction.

In this paper, we present a new method for the calculation of the Bose-Einstein correlation function with the Coulomb interaction taken into account that is applicable to a wide variety of source functions. The new technique relies on the Fourier transform of the source function and is superior to already known methods in several ways: computationally, it is much cheaper and more stable than existing methods, and conceptually it highlights the actual effect of the Coulomb interaction on the correlation function in a straightforward way. In this paper, we present the method only for spherically symmetric source functions; however, the generalization to the non-spherical case is also being worked out. Another plausible future generalization is taking more complicated final state interaction schemes, such as strong interaction, into account, along the lines of Ref. [18].

The structure of this paper is as follows. In Section 2, we detail the basic definitions and known methodology pertaining to Bose-Einstein correlations and the effect of the Coulomb interaction, paying particular attention to the role of Lévy distributions as source functions. In Section 3, we present the new method. It has some delicate mathematical intricacies that are interesting on their own (so apart from the practical benefits of our new method, its derivation might interest the purely mathematically inclined reader as well). Some mathematical details are highlighted in Section 3, and some are left to the Appendix. In Section 4, we verify our new method for the case of Lévy-stable source functions. We find that the technique is significantly helpful and augments experimental measurements by making the correlation function fitting procedure easier. We also present a software package ready to use for such measurements [19]. Section 5 summarizes our findings, pointing to possible future applications and generalizations.

Bose-Einstein correlation functions and the Coulomb interaction

The definition of the observable two-particle correlation function in a heavy-ion collision experiment is as follows:

C2(p1,p2)=N2(p1,p2)N1(p1)N1(p2)(1)C_2(\mathbf{p}_1,\mathbf{p}_2)=\frac{N_2(\mathbf{p}_1,\mathbf{p}_2)}{N_1(\mathbf{p}_1)N_1(\mathbf{p}_2)} \tag*{(1)}

where p1\mathbf{p}_1 and p2\mathbf{p}_2 are the particle momentum (three-)vectors, N1(p1)≡Ednd3pN_1(\mathbf{p}_1)\equiv E\frac{dn}{d^3p} is the single-particle invariant momentum distribution (meaning the number of particles produced within a given invariant volume element in momentum space, and EE is the particle energy corresponding to momentum p\mathbf{p}), and N2(p1,p2)≡E1E2d2nd3p1d3p2N_2(\mathbf{p}_1,\mathbf{p}_2)\equiv E_1E_2\frac{d^2n}{d^3p_1d^3p_2} is the two-particle invariant momentum distribution, i.e., the number of particle pairs produced with momenta p1\mathbf{p}_1 and p2\mathbf{p}_2 (and so with energies E1E_1 and E2E_2).

Before proceeding, let us introduce some appropriate notations and variables. Three-vectors are denoted by boldface letters, four-vectors by standard typeset letters; e.g., p1\mathbf{p}_1 is the three-momentum, p1≡(E1,p1)p_1\equiv(E_1,\mathbf{p}_1) is the four-momentum of particle 1. Lorentz products of four-vectors are denoted by a dot; for the metric tensor we use the gμν=diag⁡(1,−1,−1,−1)g^{\mu\nu}=\operatorname{diag}(1,-1,-1,-1) convention. The scalar product (Euclidean inner product) of three-vectors is denoted by writing the vectors next to each other. Below we deal with functions of two sets of space-time coordinates, x1≡(t1,r1)x_1\equiv(t_1,\mathbf{r}_1) and x2≡(t2,r2)x_2\equiv(t_2,\mathbf{r}_2); instead of these coordinates we can use the center-of-mass coordinate X≡(T,R)X\equiv(T,\mathbf{R}) and the relative coordinates x≡(t,r)x\equiv(t,\mathbf{r}) defined as

X:=12(x1+x2),T:=12(t1+t2),R:=12(r1+r2),x:=x1−x2,t:=t1−t2,r:=r1−r2.(2)\begin{aligned} X&:=\tfrac{1}{2}(x_1+x_2), &\qquad T&:=\tfrac{1}{2}(t_1+t_2), &\qquad\mathbf{R}&:=\tfrac{1}{2}(\mathbf{r}_1+\mathbf{r}_2),\\ x&:=x_1-x_2, & t&:=t_1-t_2, & \mathbf{r}&:=\mathbf{r}_1-\mathbf{r}_2. \tag*{(2)} \end{aligned}

Similarly, besides the momenta of the particles, p1p_1 and p2p_2, we use the average momentum four-vector KK, as well as the relative momentum four-vector QQ, and use the three-vector slices of these, K\mathbf{K} and Q\mathbf{Q}:

K:=12(p1+p2),Q:=p1−p2,K:=12(p1+p2),Q:=p1−p2.(3)K := \frac{1}{2}(p_1+p_2), \qquad Q := p_1-p_2, \qquad\mathbf{K} := \frac{1}{2}(\mathbf{p}_1+\mathbf{p}_2), \qquad\mathbf{Q} := \mathbf{p}_1-\mathbf{p}_2. \tag*{(3)}

Since we are dealing with identical particles, their masses are equal, m1=m2≡mm_1=m_2\equiv m. With respect to the momentum variables, we usually suppress the ℏ\hbar reduced Planck constant, meaning that momentum vectors are also understood as wave number vectors if the dimensions of the quantities imply so; for example, we write a plane wave-like wave function of the rr variable like eiQre^{iQr} but measure QQ in MeV/c like a momentum variable.

The C2(p1,p2)C_2(p_1,p_2) correlation function introduced in Eq. (1) above equals unity if the particle production is uncorrelated. In high-energy nuclear collisions, there are a number of processes that lead to C2≠1C_2\ne1, i.e., correlated particle production (collective flow, jets, resonance decays, etc.). For identical bosons (such as charged pions), the main source of correlation at low momentum difference is what is called Bose-Einstein correlation, a quantum mechanical correlation stemming from the indistinguishability of the particles. For the simplest theoretical treatment of this, one can introduce the source function S(x,p)≡S(t,r,p)S(x,p)\equiv S(t,r,p) as the probability of a particle produced at space-time point x≡(t,r)x\equiv(t,r) with momentum pp. With this, the invariant momentum distribution is readily expressed as

N1(p)=∫d4x S(x,p),(4)N_1(p)=\int d^4x\,S(x,p), \tag*{(4)}

where d4xd^4x stands for dt d3rdt\,d^3r, i.e., the integration is taken over the whole space-time; however, usually SS has a finite support or at least a rapid decrease in space and in time. Note that for the space-time coordinates, we have used tt and rr as independent variables. On the other hand, for the momentum variable, we only write out pp explicitly, with EE omitted, since for a given type of particle with mass mm, EE is already determined by pp as E=m2+p2E=\sqrt{m^2+p^2}.

Alternatively, one can argue that the Ψp(1)(x)\Psi_p^{(1)}(x) single-particle wave function describing a particle with momentum pp is a plane wave, and thus

N1(p)=∫d4x S(x,p)∣Ψp(1)(x)∣2=∫d4x S(x,p),sinceΨp(1)(x)=e−ip⋅x ⇒ ∣Ψp(1)(x)∣2=1.(5)N_1(p)=\int d^4x\,S(x,p)\left|\Psi_p^{(1)}(x)\right|^2=\int d^4x\,S(x,p), \qquad\text{since}\qquad\Psi_p^{(1)}(x)=e^{-ip\cdot x}\ \Rightarrow\ \left|\Psi_p^{(1)}(x)\right|^2=1. \tag*{(5)}

The essence of the description of the Bose-Einstein correlation lies in the statement that the analog expression of N2(p1,p2)N_2(p_1,p_2) that takes into account particle production at two distinct space-time points x1x_1 and x2x_2 must be written up utilizing the two-particle wave-function Ψp1,p2(2)(x1,x2)\Psi_{p_1,p_2}^{(2)}(x_1,x_2): [20]

N2(p1,p2)=∫d4x1 d4x2 ∣Ψp1,p2(2)(x1,x2)∣2⋅S(x1,p1)S(x2,p2),(6)N_2(p_1,p_2)=\int d^4x_1\,d^4x_2\,\left|\Psi_{p_1,p_2}^{(2)}(x_1,x_2)\right|^2\cdot S(x_1,p_1)S(x_2,p_2), \tag*{(6)}

and that for bosons, Ψp1,p2(2)(x1,x2)\Psi_{p_1,p_2}^{(2)}(x_1,x_2) is symmetric in the x1x_1 and x2x_2 variables. If the final state interactions of the particles are neglected, the wave function is a symmetrized plane wave, denoted by Ψp1,p2(2,0)\Psi_{p_1,p_2}^{(2,0)}:

Ψp1,p2(2,0)(x1,x2)=12[e−ip1⋅x1e−ip2⋅x2+e−ip1⋅x2e−ip2⋅x1].(7)\Psi_{p_1,p_2}^{(2,0)}(x_1,x_2)=\frac{1}{\sqrt{2}}\left[e^{-ip_1\cdot x_1}e^{-ip_2\cdot x_2}+e^{-ip_1\cdot x_2}e^{-ip_2\cdot x_1}\right]. \tag*{(7)}

For the modulus square of this wave function, one obtains

∣Ψp1,p2(2,0)(x1,x2)∣2=1+cos⁡(Q⋅(x1−x2)),(8)\left|\Psi_{p_1,p_2}^{(2,0)}(x_1,x_2)\right|^2=1+\cos\left(Q\cdot(x_1-x_2)\right), \tag*{(8)}

and from this, after some simplifications, one gets the following expression for the correlation function (where the 0 superscript denotes that final state interactions are neglected):

C2(0)(p1,p2)=1+Re⁡S~(Q,p1)S~∗(Q,p2)S~(0,p1)S~∗(0,p2),(9)C_2^{(0)}(p_1,p_2)=1+\operatorname{Re}\frac{\widetilde{S}(Q,p_1)\widetilde{S}^{*}(Q,p_2)}{\widetilde{S}(0,p_1)\widetilde{S}^{*}(0,p_2)}, \tag*{(9)}

with S~(Q,p)\widetilde{S}(Q,p) being the Fourier transform of the source function:

S~(Q,p)=∫d4x S(x,p)e−iQ⋅x.(10)\widetilde{S}(Q,p)=\int d^4x\,S(x,p)e^{-iQ\cdot x}. \tag*{(10)}

Typically the correlation function is written up as a function of the average and relative momentum variables KK and QQ, respectively, instead of p1p_1, p2p_2. The reason for this is that it turns out that the dependence on QQ is much stronger than on KK, so one almost always makes the so-called smoothness approximation [21] that in Eq. (9), p1≈p2≈Kp_1 \approx p_2 \approx K in the second argument of S~\tilde{S}. One then has

C2(0)(Q,K)=1+∣S~(Q,K)∣2∣S~(0,K)∣2,(11)C_2^{(0)}(Q,K)=1+\frac{\lvert\tilde{S}(Q,K)\rvert^2}{\lvert\tilde{S}(0,K)\rvert^2}, \tag*{(11)}

and one can conceptually separate the QQ and KK dependence by assuming that the source function S(x,K)S(x,K) has a given functional form (e.g., a Gaussian) as a function of xx, and the parameters of this functional form may depend on KK. By measuring the correlation function as a function of QQ at different KK values, one can reconstruct the KK-dependence of the source parameters.

We can express the Ψ(2,0)\Psi^{(2,0)} two-particle wave function in terms of the center-of-mass system variables and the relative variables as

ΨQ,K(2,0)(x,X)=e−2iK⋅X⋅ψQ,K(0)(x),(12)\Psi_{Q,K}^{(2,0)}(x,X)=e^{-2iK\cdot X}\cdot\psi_{Q,K}^{(0)}(x), \tag*{(12)}

where

ψQ,K(0)(x)=12[e−i2Q⋅x+ei2Q⋅x].(13)\psi_{Q,K}^{(0)}(x)=\frac{1}{\sqrt{2}}\left[e^{-\frac{i}{2}Q\cdot x}+e^{\frac{i}{2}Q\cdot x}\right]. \tag*{(13)}

In this form, we see that the relative motion and the center-of-mass motion disentangle. In the case of non-vanishing final state interactions, the center-of-mass motion still retains its plane-wave e−2iQXe^{-2iQX} form. However, the form of the wave function pertaining to the relative motion changes. In this case, the two-particle wave function (now with the 0 superscript omitted, owing to the presence of final state interaction) is

ΨQ,K(2)(x,X)=e−2iK⋅X⋅ψQ,K(x)⇒∣ΨQ,K(2)(x,X)∣2=∣ψQ,K(x)∣2.(14)\Psi_{Q,K}^{(2)}(x,X)=e^{-2iK\cdot X}\cdot\psi_{Q,K}(x) \quad\Rightarrow\quad \lvert\Psi_{Q,K}^{(2)}(x,X)\rvert^2=\lvert\psi_{Q,K}(x)\rvert^2. \tag*{(14)}

Here the ψQ,K(x)\psi_{Q,K}(x) part of the wave function describes the relative motion. It is symmetric to the x↔−xx\leftrightarrow-x exchange (corresponding to the x1↔x2x_1\leftrightarrow x_2 change), and its concrete form depends on the interaction. We see that only this function determines the object of interest, the modulus of the two-particle wave function: the plane wave corresponding to the free motion of the center of mass cancels.

With this notation, we may rewrite the expression of the two-particle momentum distribution and the (utilizing the smoothness approximation, p1≈p2≈Kp_1\approx p_2\approx K, just as above) as

N2(Q,K)=∫d4x d4X S(X+12x1,K)S(X−12x,K)∣ψQ,K(x)∣2,(15)N_2(Q,K)=\int d^4x\,d^4X\,S\left(X+\frac{1}{2}x_1,K\right)S\left(X-\frac{1}{2}x,K\right)\lvert\psi_{Q,K}(x)\rvert^2, \tag*{(15)}
N1(K)N1(K)=∫d4x d4X S(X+12x1,K)S(X−12x,K).(16)N_1(K)N_1(K)=\int d^4x\,d^4X\,S\left(X+\frac{1}{2}x_1,K\right)S\left(X-\frac{1}{2}x,K\right). \tag*{(16)}

So with a new definition, we have

C2(Q,K)=∫d4x D(x,K)∣ψQ,K(x)∣2∫d4x D(x,K),(17)C_2(Q,K)=\frac{\int d^4x\,D(x,K)\lvert\psi_{Q,K}(x)\rvert^2}{\int d^4x\,D(x,K)}, \tag*{(17)}

where we introduced the D(x,K)D(x,K) relative coordinate distribution or spatial correlation function as

D(r,K)≡∫d4X S(X+12x,K)S(X−12x,K).(18)D(r,K)\equiv\int d^4X\,S\left(X+\frac{1}{2}x,K\right)S\left(X-\frac{1}{2}x,K\right). \tag*{(18)}

Note that while S(x,p)S(x,p) is not necessarily an even function of xx, this D(x,K)D(x,K) function is even in xx. Taking this into account, and that from Eq. (13), we have ∣ψQ,K(0)(x)∣2=1+cos⁡(Q⋅x)\lvert\psi_{Q,K}^{(0)}(x)\rvert^2=1+\cos(Q\cdot x), we can write the correlation function in the case of no final state interaction as

C2(0)(Q,K)=1+D~(Q,K)D~(0,K),whereD~(Q,K)=∫d4x D(x,K)e−iQ⋅x.(19)C_2^{(0)}(Q,K)=1+\frac{\tilde{D}(Q,K)}{\tilde{D}(0,K)}, \qquad \text{where}\qquad \tilde{D}(Q,K)=\int d^4x\,D(x,K)e^{-iQ\cdot x}. \tag*{(19)}

We see that owing to the presence of the denominator in Eqs. (17) and (19), the overall normalization of DD (the integral over space-time, at any given K\mathbf{K}) cancels from the measurable C2C_2 correlation function. To simplify our formulas, in the following, we thus assume that DD is normalized to unity at any given K\mathbf{K}.

For actual calculations, it would be desirable to have the time variable (in an appropriate reference frame) canceled from the production described by S(x,p)S(x,p). The freeze-out duration Δτ\Delta\tau is in principle not zero (indeed, a prime goal of HBT studies is to measure Δτ\Delta\tau, and thus the order of phase transition). However, as multiple calculations show (see, e.g., Ref. [22] for an early exposition), the effect of Δτ\Delta\tau being non-zero can essentially be factored into the geometric (space-like) radii of the source function (with the time component of the relative momentum, q0q^0 being expressed with the spatial components along the way). So in the following, we write our formulas as if D(x,K)D(x,K) was non-zero only if t=0t=0, highlighting this by omitting the time-dependence in the notation of our quantities, and remember that this is understood in the sense that the effect of finite time particle emission was treated by incorporating its effects into the geometrical radii of the source function. We thus have

C2(Q,K)=∫d3r D(r,K)∣ψQ,K(r)∣2,(20)C_2(\mathbf{Q},\mathbf{K})=\int d^3r\,D(\mathbf{r},\mathbf{K})|\psi_{\mathbf{Q},\mathbf{K}}(\mathbf{r})|^2, \tag*{(20)}

specifically, for the case when final state interactions are neglected,

C2(0)(Q,K)=1+∫d3r D(r,K)eiQr.(21)C_2^{(0)}(\mathbf{Q},\mathbf{K})=1+\int d^3r\,D(\mathbf{r},\mathbf{K})e^{i\mathbf{Q}\mathbf{r}}. \tag*{(21)}

We also appropriately change our notation at this point: since we no longer need four-vectors, in the case of a three-vector (denoted by bold letter), the same letter in standard typeset indicates the magnitude of the vector, e.g., Q≡∣Q∣Q\equiv|\mathbf{Q}|, r≡∣r∣r\equiv|\mathbf{r}|.

In the case of the Coulomb interaction, the ψ(x)\psi(x) wave function (a solution to the Schrödinger equation satisfying appropriate boundary conditions) has a known form in the non-relativistic case. This implies that one has to use the center-of-mass frame of the particle pair (PCMS frame), where their motion can be approximated to be non-relativistic. In the following thus, we take Q\mathbf{Q} to be understood in this PCMS frame and use the k=Q/2k=Q/2 notation, and (as customary for such treatment of the Coulomb interaction) assume that the simultaneity condition expounded above holds in PCMS. So, in this case, we have

C2(Q)=∫d3r D(r)∣ψk(r)∣2,(22)C_2(Q)=\int d^3r\,D(r)|\psi_k(r)|^2, \tag*{(22)}

where we dropped the KK arguments for simplicity, but remember that the source function depends on KK through the KK-dependence of its assumed parameters, and use the appropriately symmetrized ψk(r)\psi_k(r) Coulomb interacting out state in the following form, well known in quantum mechanical scattering theory (see, e.g., Ref. [23] for details):

ψk(r)=N∗2e−ikr[M(1−iη,1,i(kr+kr))+M(1−iη,1,i(kr−kr))].(23)\psi_k(r)=\frac{\mathcal{N}^*}{\sqrt{2}}e^{-ikr}\left[M(1-i\eta,1,i(kr+\mathbf{k}\mathbf{r}))+M(1-i\eta,1,i(kr-\mathbf{k}\mathbf{r}))\right]. \tag*{(23)}

The η\eta quantity appearing here is the so-called Sommerfeld parameter:

η≔mc2α2ℏkc,(24)\eta\coloneqq\frac{mc^2\alpha}{2\hbar kc}, \tag*{(24)}

with α≡qe24πϵ01ℏc≈1137\alpha\equiv\frac{q_e^2}{4\pi\epsilon_0}\frac{1}{\hbar c}\approx\frac{1}{137} being the fine structure constant. The quantity N\mathcal{N} is the normalization factor of the wave function, of whose square modulus is the so-called Gamow factor:

N=e−πη/2Γ(1+iη)⇒∣N∣2=2πηe2πη−1,(25)\mathcal{N}=e^{-\pi\eta/2}\Gamma(1+i\eta) \quad\Rightarrow\quad |\mathcal{N}|^2=\frac{2\pi\eta}{e^{2\pi\eta}-1}, \tag*{(25)}

where Γ(z)\Gamma(z) is the Gamma function, and its simple properties (such as zΓ(z)=Γ(z+1)z\Gamma(z)=\Gamma(z+1) and Γ(z)Γ(1−z)=πsin⁡(πz)\Gamma(z)\Gamma(1-z)=\frac{\pi}{\sin(\pi z)}) lead to the expression written up for ∣N∣2|\mathcal{N}|^2. Finally, M(a,b,z)M(a,b,z) is the confluent hypergeometric function,

M(a,b,z)=1+abz1!+a(a+1)b(b+1)z22!+…=∑n=0∞Γ(a+n)Γ(a)Γ(b)Γ(b+n)znn!.(26)M(a,b,z)=1+\frac{a}{b}\frac{z}{1!}+\frac{a(a+1)}{b(b+1)}\frac{z^2}{2!}+\ldots=\sum_{n=0}^{\infty}\frac{\Gamma(a+n)}{\Gamma(a)}\frac{\Gamma(b)}{\Gamma(b+n)}\frac{z^n}{n!}. \tag*{(26)}

Besides the (modulus square of the) wave function, an assumption is needed about the functional form of the D(r)D(\mathbf{r}) pair distribution in order to calculate the correlation function as in Eq. (22). As mentioned in § 1, recent measurements utilized Lévy-stable distributions as an assumption for D(r)D(\mathbf{r}). The appearance of Lévy-stable sources in high energy heavy-ion collisions was expounded in [24]. Since then, several mechanisms were proposed as the possible cause for this particular source shape, such as jet fragmentation [25], critical phenomena [26], directional averaging and non-sphericality [27], event averaging [27], resonance decays and hadronic rescattering [28, 29]. Out of these, each may result in Lévy-shaped sources at different collision energies and systems: jet fragmentation in e+e−e^{+}e^{-} or pp collisions, critical phenomena at energies lower than the top RHIC or LHC energies, while hadronic rescattering may occur generally. In [29], it was shown that in realistic simulations of the hadronic medium, Lévy sources appear on an event-by-event basis at various stages of the evolution.

The symmetric Lévy distribution is a generalization of the Gaussian function as it has an additional parameter α\alpha, called the Lévy exponent. In the spherically symmetric case, it has the following form:

L(α,R,r):=1(2π)3∫d3q eiqre−12∣qR∣α.(27)\mathcal{L}(\alpha,R,\mathbf{r}) := \frac{1}{(2\pi)^3}\int\mathrm{d}^3q\, e^{i\mathbf{q}\mathbf{r}}e^{-\frac{1}{2}\lvert qR\rvert^\alpha}. \tag*{(27)}

By the known property of Fourier transforms, one sees that this function is normalized to unity:

∫d3r L(α,R,r)=1.(28)\int\mathrm{d}^3r\,\mathcal{L}(\alpha,R,\mathbf{r})=1. \tag*{(28)}

In the α=2\alpha=2 case we recover a Gaussian distribution with radius RR; while for α=1\alpha=1, a Cauchy distribution is obtained:

L(α=2,R,r)=1(2πR2)3/2e−r22R2,L(α=1,R,r)=8π2R31(1+4r2R2)2.(29)\mathcal{L}(\alpha=2,R,\mathbf{r})=\frac{1}{(2\pi R^2)^{3/2}}e^{-\frac{r^2}{2R^2}},\qquad\mathcal{L}(\alpha=1,R,\mathbf{r})=\frac{8}{\pi^2R^3}\frac{1}{\left(1+\frac{4r^2}{R^2}\right)^2}. \tag*{(29)}

For α≠2\alpha\ne2 and α≠1\alpha\ne1, one cannot express L(α,R,r)\mathcal{L}(\alpha,R,\mathbf{r}) with a combination of simple functions. However, its asymptotic expression is known: for every α≠2\alpha\ne2, it decreases like a power of rr:

for large r,r2L(α,R,r)∝r−1−α,(30)\text{for large }r,\qquad r^2\mathcal{L}(\alpha,R,\mathbf{r})\propto r^{-1-\alpha}, \tag*{(30)}

as exemplified in the above α=1\alpha=1 case as well.

Lévy distributions generalize Gaussian distributions also in the sense that they retain the stability property of Gaussian: the convolution of two Lévy distributions (with the same α\alpha parameter) is again such a Lévy function, as can be directly seen by inserting the above definition (a Fourier integral) and performing the integrals. Specifically,

∫d3ρ L(α,R1,ρ+12r)L(α,R2,ρ−12r)=L(α,R,r),withRα=R1α+R2α.(31)\int\mathrm{d}^3\rho\,\mathcal{L}\left(\alpha,R_1,\boldsymbol{\rho}+\frac{1}{2}\mathbf{r}\right)\mathcal{L}\left(\alpha,R_2,\boldsymbol{\rho}-\frac{1}{2}\mathbf{r}\right)=\mathcal{L}(\alpha,R,\mathbf{r}),\qquad\text{with}\quad R^\alpha=R_1^\alpha+R_2^\alpha. \tag*{(31)}

From another point of view, with a similar calculation, we obtain that if the S(r,p)S(\mathbf{r},p) single-particle source function is a Lévy-type function, then so is the D(r,p)D(\mathbf{r},p) two-particle source function:

S(r,p)=L(α(p),R(p),r−r0(p))⇒D(r,p)=L(α(p),21/α(p)R(p),r).(32)S(\mathbf{r},p)=\mathcal{L}\left(\alpha(p),R(p),\mathbf{r}-\mathbf{r}_0(p)\right)\quad\Rightarrow\quad D(\mathbf{r},p)=\mathcal{L}\left(\alpha(p),2^{1/\alpha(p)}R(p),\mathbf{r}\right). \tag*{(32)}

We could even allow pp-dependent parameters as well as an arbitrary r0(p)\mathbf{r}_0(p) shift in S(r,p)S(\mathbf{r},p); this does not change the shape of D(r,p)D(\mathbf{r},p), the quantity that appears in the expression of the correlation function.

The stability property written up in Eq. (31) is the main reason why Lévy distributions are expected to appear in many different circumstances (as limiting distributions, just as Gaussians). In the scenarios mentioned above, they also naturally arise as particle production source functions in heavy-ion collisions, the main object of interest in this paper. Indeed, Lévy distributions have already been successfully applied to describe such sources [15], after earlier measurements indicated that a Gaussian description falls short of such data [13]. However, as explained below, it turns out that the actual calculation of the C2C_2 correlation function for Lévy-like D(r,p)D(\mathbf{r},p) is cumbersome even numerically; there is thus a natural need for the development of such calculational methods. This becomes even more pressing if one abandons spherical symmetry or if one allows further generalizations of the source function beyond the Lévy assumption.

A new method for the treatment of the Coulomb effect

As written up in Eq. (22) above, for calculating the correlation function with the Coulomb interaction taken into account, one has to integrate the modulus of the Coulomb interacting wave function ψk(r)\psi_k(r) weighted with the D(r)D(r) two-particle source function. This is deemed to be feasible only numerically, owing to the complicated form of ψk(r)\psi_k(r), Eq. (23). Nevertheless, as seen above in Eq. (21), in case of vanishing Coulomb force, ∣ψk(r)∣|\psi_k(r)| is a plane wave, and our desired integral reduces to a Fourier transformation of D(r)D(r).

There is an important, if not the only interesting, class of possible source function assumptions that have the property of being defined and best calculable as a Fourier transform of some simple function. A Gaussian source function obviously falls into this category; however, a preeminent motivating example is when the D(r)D(r) two-particle source function is a Lévy distribution, whose Fourier transform, e−∣Rq∣αe^{-|Rq|^\alpha} is an easily computable fast decreasing function. In the following, we thus assume that

D(r)=1(2π)3∫d3q f(q)eiqr.(33)D(\mathbf{r}) = \frac{1}{(2\pi)^3}\int\mathrm{d}^3q\,f(\mathbf{q})e^{i\mathbf{q}\mathbf{r}}. \tag*{(33)}

We assume that f(q)f(q) is an integrable function (over the whole qq space); this is a natural assumption if we want to interpret this Fourier transform as a regular (Lebesgue-)integral. We also assume D(r)D(r) to be an integrable function (over the whole rr space; this is naturally fulfilled whenever S(r)S(r) is integrable, which is necessary to interpret S(r)S(r) as the function whose integral, according to Eq. (4), gives the single particle invariant momentum distribution). These imply that both f(q)f(q) and D(r)D(r) are bounded and continuous, as well as that the (inverse) Fourier transform of D(q)D(q) is also a regular integral:

f(q)=∫d3r D(r)e−iqr,(34)f(q)=\int\mathrm{d}^3r\,D(r)e^{-iqr}, \tag*{(34)}

and thus the normalization condition for D(r)D(r) means a simple condition on ff:

∫d3r D(r)=1⇔f(q=0)=1.(35)\int\mathrm{d}^3r\,D(r)=1 \qquad\Leftrightarrow\qquad f(q=0)=1. \tag*{(35)}

In particular, the interaction-free correlation function has ff as the main component; from Eqs. (21) and (34),

C2(0)(Q)=1+f(Q).(36)C_2^{(0)}(Q)=1+f(Q). \tag*{(36)}

In the Coulomb interacting case, our goal is to calculate the correlation function using Eq. (22); put together with Eq. (33), we have

C2(Q)=1(2π)3∫d3r ∣ψk(r)∣2∫d3q f(q)eiqr,(37)C_2(Q)=\frac{1}{(2\pi)^3}\int\mathrm{d}^3r\,|\psi_k(r)|^2\int\mathrm{d}^3q\,f(q)e^{i\mathbf{q}\mathbf{r}}, \tag*{(37)}

where ψk(r)\psi_k(r) is the Coulomb wave function written up in Eq. (23). The straightforward way to proceed used, e.g., in [18, 30] in the case of Lévy distributions, is then to calculate D(r)D(r) from f(q)f(q) as a Fourier integral, then perform the integral over rr, in which ∣ψk(r)∣2|\psi_k(r)|^2 enters. Numerically, this is a daunting task in some cases. E.g., for the Lévy distribution, the result of the Fourier transform (the function D(r)D(r) itself) is only slowly decreasing (with a power-law-like behavior for large rr), while ∣ψk(r)∣2|\psi_k(r)|^2 is an oscillating function; asymptotically a plane wave (up to logarithmic corrections). There is also a conceptual awkwardness in this methodology: we take a Fourier transform of f(q)f(q) and then subject the resulting D(r)D(r) function to an “almost inverse Fourier transform” (i.e., the integral with the “almost plane wave” ∣ψk(r)∣2|\psi_k(r)|^2 as a kernel) to finally arrive at C2(Q)C_2(Q). In the interaction-free case, ∣ψk(r)∣2|\psi_k(r)|^2 is really a plane wave, and the result, Eq. (36) indeed shows that these two Fourier transformations cancel each other. One would thus very much prefer a calculational scheme where this “back and forth” transformation is not needed in its full numerical complexity. (For example, calculating C2(Q)C_2(Q) at high QQ values, where owing to the decrease of f(q)f(q) it is increasingly closer to unity, still requires significant computational power because at higher kk, the oscillations of ψk(r)\psi_k(r) become faster and faster.)

The natural idea to resolve these problems is that we would like to perform the Fourier transform of the interacting ∣ψk(r)∣2|\psi_k(r)|^2 function and “act” with this resulting integral kernel (a function of kk and qq) on the f(q)f(q) function. In other words, we would like to interchange the order of integrals in Eq. (37) so that we could write

C2(Q)=?1(2π)3∫d3q f(q)∫d3r eiqr∣ψk(r)∣2.(38)C_2(Q)\mathrel{\overset{?}{=}}\frac{1}{(2\pi)^3}\int\mathrm{d}^3q\,f(q)\int\mathrm{d}^3r\,e^{i\mathbf{q}\mathbf{r}}|\psi_k(r)|^2. \tag*{(38)}

Regrettably, however, this is not possible in such a simple form: since ψk(r)\psi_k(r) is asymptotically a plane wave, its modulus is definitely not an integrable function whose Fourier transform could simply be calculated by an integral. In the following, we work around this problem by carefully treating integrability and the interchange of integrals and limits, a rarely found exercise in physics-motivated standard calculations. The resulting formulas, however, are worth such careful investigation. In doing so, we utilize some fundamental theorems about (Lebesgue) integrability.3 We outline the main steps of our calculations below but leave some mathematical and calculational details to the Appendix.

It turns out that we can proceed by inserting an exponential “regularization”, e−λre^{-\lambda r} into the integrand of Eq. (37), with a positive real λ>0\lambda> 0 parameter, which at the end goes to 0. With finite λ\lambda we can interchange the order of integrals and finally arrive at

C2(Q)=∫d3r ∣ψk(r)∣2∫d3q(2π)3f(q)eiqr=∫d3r lim⁡λ→0e−λr∣ψk(r)∣2∫d3q(2π)3f(q)eiqr=1=1lim⁡λ→0∫d3r∫d3q(2π)3e−λr∣ψk(r)∣2f(q)eiqr=2=2lim⁡λ→0∫d3q(2π)3f(q)∫d3r e−λr∣ψk(r)∣2eiqr.(39)\begin{aligned} C_2(Q) &= \int\mathrm{d}^3r\,|\psi_k(r)|^2 \int\frac{\mathrm{d}^3q}{(2\pi)^3}f(q)e^{i\mathbf{q}\mathbf{r}} \\ &= \int\mathrm{d}^3r\,\lim_{\lambda\to0}e^{-\lambda r}|\psi_k(r)|^2\int\frac{\mathrm{d}^3q}{(2\pi)^3}f(q)e^{i\mathbf{q}\mathbf{r}} \overset{1}{=} \\ &\overset{1}{=} \lim_{\lambda\to0}\int\mathrm{d}^3r\int\frac{\mathrm{d}^3q}{(2\pi)^3}e^{-\lambda r}|\psi_k(r)|^2f(q)e^{i\mathbf{q}\mathbf{r}} \overset{2}{=} \\ &\overset{2}{=} \lim_{\lambda\to0}\int\frac{\mathrm{d}^3q}{(2\pi)^3}f(q)\int\mathrm{d}^3r\,e^{-\lambda r}|\psi_k(r)|^2e^{i\mathbf{q}\mathbf{r}}. \tag*{(39)} \end{aligned}

(Remember that QQ is the variable of the resulting observable C2C_2 correlation function, Q=2kQ = 2k, while qq is a mere auxiliary integration variable in this calculation.)

In Eq. (39), every step is justified mathematically, in particular, the exchange of the limit and the integral in Step 1 (by virtue of Lebesgue’s theorem), and the exchange of the order of integrals in Step 2 (by virtue of Fubini’s theorem); see Appendix A.1 for details. Nevertheless, the point is that in the resulting final formula, the λ→0\lambda\to0 limit cannot be exchanged with the qq-integral. The way to proceed is to calculate the rr-integral for finite λ\lambda values, and then “simplify” the λ→0\lambda\to0 limit to arrive at a final formula where there are no explicit limits that would have to be evaluated numerically (which would be very challenging if not impossible).

In this paper, we proceed with pair source functions that are spherically symmetric; this also implies spherical symmetry of f(q)f(q), which we highlight everywhere with an ss in the subscript. In the spherically symmetric case, we can perform the solid angle integral in the Fourier integral, arriving at the following expression of D(r)D(r):

f(q)≡fs(q)⇒D(r)≡D(r)=12π2r∫0∞dq qsin⁡(qr)fs(q).(40)f(q) \equiv f_s(q) \quad\Rightarrow\quad D(r) \equiv D(r) = \frac{1}{2\pi^2r}\int_0^\infty\mathrm{d}q\,q\sin(qr)f_s(q). \tag*{(40)}

Following the exact same steps as in Eq. (39), with this expression for the spherically symmetric D(r)D(r) we have

C2(Q)=12π2lim⁡λ→0∫0∞dq q2fs(q)∫d3r e−λrsin⁡(qr)qr∣ψk(r)∣2.(41)C_2(Q)=\frac{1}{2\pi^2}\lim_{\lambda\to0}\int_0^\infty\mathrm{d}q\,q^2f_s(q)\int\mathrm{d}^3r\,e^{-\lambda r}\frac{\sin(qr)}{qr}|\psi_k(r)|^2. \tag*{(41)}

Note that while ψk(r)\psi_k(r) depends on the direction of k=Q/2\mathbf{k} = \mathbf{Q}/2, the result of this integral indeed only depends on the magnitude QQ, after performing the rr-integral.

As a next step, we substitute ψk(r)\psi_k(r) from Eq. (23). The modulus square results in four terms (from each pairings of the different M(a,b,z)M(a,b,z) functions). With an appropriate r→−r\mathbf{r} \to-\mathbf{r} change of variable in two of these, we are left with only two different terms:

C2(Q)=∣N∣22π2lim⁡λ→0∫0∞dq q2fs(q)[D1λ(q)+D2λ(q)],where(42)C_2(Q)=\frac{|\mathcal{N}|^2}{2\pi^2}\lim_{\lambda\to0}\int_0^\infty\mathrm{d}q\,q^2f_s(q)\left[D_{1\lambda}(q)+D_{2\lambda}(q)\right], \quad\text{where} \tag*{(42)}
D1λ(q)=∫d3r sin⁡(qr)qre−λrM(1+iη,1,−i(kr+kr))M(1−iη,1,i(kr+kr)).(43)D_{1\lambda}(q)=\int\mathrm{d}^3r\,\frac{\sin(qr)}{qr}e^{-\lambda r}M(1+i\eta,1,-i(kr+\mathbf{k}\mathbf{r}))M(1-i\eta,1,i(kr+\mathbf{k}\mathbf{r})). \tag*{(43)}
D2λ(q)=∫d3r sin⁡(qr)qre−λrM(1+iη,1,−i(kr−kr))M(1−iη,1,i(kr−kr)).(44)D_{2\lambda}(q)=\int\mathrm{d}^3r\,\frac{\sin(qr)}{qr}e^{-\lambda r}M(1+i\eta,1,-i(kr-\mathbf{k}\mathbf{r}))M(1-i\eta,1,i(kr-\mathbf{k}\mathbf{r})). \tag*{(44)}

The integrals of Eqs. (43) and (44) are calculated in Appendix A.2; the utilized method is similar to that of Nordsieck [31]. In our case, the result is

D1λs(q)=4πqIm⁡[1(λ−iq)2(1+2kq+iλ)2iη2F1(iη,1+iη,1,4k2(q+iλ)2)](45)\mathcal{D}_{1\lambda s}(q)=\frac{4\pi}{q}\operatorname{Im}\left[\frac{1}{(\lambda-iq)^2}\left(1+\frac{2k}{q+i\lambda}\right)^{2i\eta}{}_2F_1\left(i\eta,1+i\eta,1,\frac{4k^2}{(q+i\lambda)^2}\right)\right] \tag*{(45)}
D2λs(q)=4πqIm⁡[(λ−iq−2ik)iη(λ−iq+2ik)−iη(λ−iq)2+4k2](46)\mathcal{D}_{2\lambda s}(q)=\frac{4\pi}{q}\operatorname{Im}\left[\frac{(\lambda-iq-2ik)^{i\eta}(\lambda-iq+2ik)^{-i\eta}}{(\lambda-iq)^2+4k^2}\right] \tag*{(46)}

where we make use of the (ordinary) hypergeometric function:

2F1(a,b,c,z)=1+abcz1!+a(a+1)b(b+1)c(c+1)z22!+…=∑n=0∞Γ(a+n)Γ(b+n)Γ(a)Γ(b)Γ(c)Γ(c+n)znn!.(47){}_2F_1(a,b,c,z)=1+\frac{ab}{c}\frac{z}{1!}+\frac{a(a+1)b(b+1)}{c(c+1)}\frac{z^2}{2!}+\ldots=\sum_{n=0}^{\infty}\frac{\Gamma(a+n)\Gamma(b+n)}{\Gamma(a)\Gamma(b)}\frac{\Gamma(c)}{\Gamma(c+n)}\frac{z^n}{n!}. \tag*{(47)}

This power series is valid only for ∣z∣<1\lvert z\rvert<1, however, the domain of 2F1(a,b,c,z){}_2F_1(a,b,c,z) can be extended to any zz complex number arguments by analytic continuation; see e.g. Ref. [32].

The next step is then to try to evaluate C2(Q)C_2(Q) as in Eq. (42) from these expressions of D1λs\mathcal{D}_{1\lambda s} and D2λs\mathcal{D}_{2\lambda s}. For this, we need to take the λ→0\lambda\to0 limit in a careful way. It turns out that as λ\lambda approaches 0, the functional forms of D1λs\mathcal{D}_{1\lambda s} and D2λs\mathcal{D}_{2\lambda s} become ill-behaved; the desired limit (of the result of the qq-integral) cannot simply be expressed as some limiting qq-function multiplied by fs(q)f_s(q) and then integrated. Instead, in the final expression, one has to take a specific linear functional of fs(q)f_s(q). 4 With leaving some intermediate steps to Appendix A.3, in the following Section, we begin by writing up the main result of our calculation for C2(Q)C_2(Q).

Results and discussion

The new formula for C2(Q)C_2(Q)

Starting from the expression of C2(Q)C_2(Q) from Eq. (42), and substituting D1λs\mathcal{D}_{1\lambda s} and D2λs\mathcal{D}_{2\lambda s} from Eqs. (45) and (46), a careful utilization of Lebesgue’s theorem (detailed in Appendix A.3) allows one to perform the remaining λ→0\lambda\to0 limit in Eq. (42). We thus arrive at the main result of our new method for calculating the Coulomb interacting C2(Q)C_2(Q), which can be summarized as follows:

C2(Q)=∣N∣2(1+fs(2k)+ηπ[A1s+A2s])(48)C_2(Q)=\lvert\mathcal{N}\rvert^2\left(1+f_s(2k)+\frac{\eta}{\pi}\left[A_{1s}+A_{2s}\right]\right) \tag*{(48)}

where the A1sA_{1s} and A2sA_{2s} terms are the following functionals of fs(q)f_s(q):

A1s=−2η∫0∞dq fs(q)−fs(0)qIm⁡[(1+2kq)2iη2F1(iη,1+iη,1,4k2q2−i0)](49)A_{1s}=-\frac{2}{\eta}\int_0^\infty dq\,\frac{f_s(q)-f_s(0)}{q}\operatorname{Im}\left[\left(1+\frac{2k}{q}\right)^{2i\eta}{}_2F_1\left(i\eta,1+i\eta,1,\frac{4k^2}{q^2}-i0\right)\right] \tag*{(49)}
A2s=−2η∫0∞dq fs(q)−fs(2k)q−2kqq+2kIm⁡(q+2k)iη(q−2k+i0)iη.(50)A_{2s}=-\frac{2}{\eta}\int_0^\infty dq\,\frac{f_s(q)-f_s(2k)}{q-2k}\frac{q}{q+2k}\operatorname{Im}\frac{(q+2k)^{i\eta}}{(q-2k+i0)^{i\eta}}. \tag*{(50)}

If η→0\eta\to0 (either by letting k→∞k\to\infty, or by formally turning off the Coulomb force), A1sA_{1s} and A2sA_{2s} take on finite values (because both the η\eta denominator and the imaginary values of the denoted quantities go to 0). So for η→0\eta\to0, in Eq. (48) the contributions of A1sA_{1s} and A2sA_{2s} vanish (where they are again multiplied by η\eta); we wrote the η\eta factors in the way we did to highlight this feature.

Note also that in Eq. (48), the effect of Coulomb interaction enters in a straightforward, “traceable” way. Omitting ∣N∣2\lvert\mathcal{N}\rvert^2 as well as the A1sA_{1s}, A2sA_{2s} terms, and recalling that Q=2kQ=2k, we simply have Eq. (36), the result for the interaction-free case. Including the ∣N∣2\lvert\mathcal{N}\rvert^2 factor (but not A1sA_{1s} and A2sA_{2s}) is the so-called Gamow correction, the

lim⁡λ→0∫−∞∞dxπλλ2+(x−x0)2f(x)=f(x0).\lim_{\lambda\to0}\int_{-\infty}^{\infty}\frac{dx}{\pi}\frac{\lambda}{\lambda^2+(x-x_0)^2}f(x)=f(x_0).

Based on this, one can say that the narrower and higher Lorentz curves “approximate the δ(x−x0)\delta(x-x_0) delta function”; the meaning of this statement is essentially the same as the displayed identity. From a practical point of view, the benefit of this formula is that instead of numerically performing some λ→0\lambda\to0 limit of the result of the left-hand side integral, the right-hand side provides a much simpler calculational statement (by requiring just to evaluate the ff function at a given x0x_0 point); however, this simpler statement is no longer a true integral transformation acting on ff.

simplest approximate treatment of the Coulomb interaction: it treats the source as point-like for the Coulomb integration (but not for the calculation of the interaction-free correlation function). The A1sA_{1s} and A2sA_{2s} terms can thus be thought of as the effect of the source being not point-like; the parts of the correlation function that is neglected by the Gamow prescription.

In our normalization, fs(0)=1f_s(0)=1; however, we retained fs(0)f_s(0) in Eq. (49) to highlight that the fraction containing the fsf_s function is a well-defined function even around q=0q=0 (provided that fsf_s is continuously differentiable). Similarly, the fraction containing fsf_s in the expression of A2sA_{2s}, Eq. (50), is also a well-defined function for continuously differentiable fsf_s, even around q=2kq=2k. In fact, we can loosen the requirement of fsf_s being continuously differentiable. The more general assumption under which the above formulas are derived is explained (along with other details) in Appendix A.3 around Eqs. (82) and (83). Here we just remark that this more general assumption is satisfied already if f(q)f(q) is everywhere continuously differentiable, and also in case of Lévy distributions (for which f(q)f(q) is not differentiable at the origin if α≤1\alpha\le1).

Power functions of complex variables are understood in the strictly single-valued function sense: for z,w∈Cz,w \in\mathbb{C}, zw:=exp⁡(wLn⁡z)z^w := \exp(w\operatorname{Ln}z), where exp⁡\exp is an entire function, its inverse however, Ln⁡z≡log⁡(∣z∣)+iarg⁡z\operatorname{Ln}z \equiv\log(\lvert z\rvert)+i\arg z, has a branch cut along the R0−\mathbb{R}_{0}^{-} negative real line (owing to the jump of arg⁡z\arg z, the phase of zz, from π\pi to −π-\pi, when crossing this line in a counterclockwise direction). This branch cut along R0−\mathbb{R}_{0}^{-} is inherited by the power function zwz^w as a function of zz, whenever w∉Zw \notin\mathbb{Z}. So when taking a power of a quantity that is on this branch cut, we have to specify which side of the cut we are on. This is the reason for the +i0+i0 term in Eq. (50). Also when one defines the 2F1(a,b,c,z)_2F_1(a,b,c,z) hypergeometric function for ∣z∣≥1\lvert z\rvert\ge1, that is, outside the convergence radius of the power series in Eq. (47), one encounters a branching point at z=1z=1, and the usual convention places the branch cut on the [1,∞]⊂R+[1,\infty] \subset\mathbb{R}^{+} real line. The −i0-i0 term in Eq. (49) specifies which side of this cut we are on.

Utilizing the new formula for Lévy-stable sources

Before delving into the details of the applied computational methods, in Figure 1, we show the main results of our calculations, i.e., the C2(Q)C_2(Q) correlation functions calculated for Lévy-stable source distributions L(α,R,r)L(\alpha,R,r), separately for pion-pion and kaon-kaon pairs. Different colors denote different Lévy scale RR values, while the dashed and solid lines correspond to the Lévy exponent α\alpha values of 0.6 and 2, respectively. Correlation functions with α\alpha between these two extremal values are illustrated with a filled area. In the interaction-free case, for any given RR, correlation functions take the same value for any α\alpha at one specific Q>0Q>0 value: C2(0)(Q=ℏc/R)=1+1/eC_2^{(0)}(Q=\hbar c/R)=1+1/e. In the case when Coulomb interaction is included, the curves with different α\alpha values still intersect each other at approximately the same point (for a given RR), in the vicinity of Q≈ℏc/RQ\approx\hbar c/R. However, in this case, the C2(Q≈ℏc/R)C_2(Q\approx\hbar c/R) values have an RR dependence. Moreover, the intersection of all the curves is only approximate (albeit a very good one); higher zooming around the apparent “nodes” would tell the curves apart.

Figure 1. Example correlation functions for pions (left) and kaons (right), plotted for four different RR and two α\alpha values. At a given RR value, the shape of the correlation function with increasing α\alpha values from α=0.6\alpha=0.6 to α=2\alpha=2 goes smoothly through the shaded region.

Testing the final numerical integral

As described above in Eq. (48), for the correlation function, two numerical integrals have to be performed. For a reasonable fs(q)f_s(q) function, such as the one for Lévy distributions, the integrands there are particularly well-behaving (non-oscillating, smooth) functions. The integrals over q∈R+0q \in\mathbb{R}_{+}^{0} are best transformed to integrals over the [0,1][0,1] interval, with a new integration variable xx introduced separately on the q∈[0,2k]q \in[0,2k] and the q∈[2k,∞]q \in[2k,\infty] intervals as q=2kxq=2kx, and q=2kxq=\frac{2k}{x}, respectively. The integrands turn out to be well-behaving „tame” functions of this new xx variable; they are written up as

A1s=−2η∫01dx Im⁡[fs(2kx)−fs(0)x(1+1x)2iη2F1(iη,1+iη,1,1x2−i0)+fs(2kx)−fs(0)x(1+x)2iη2F1(iη,1+iη,1,x2−i0)].(51)A_{1s}=-\frac{2}{\eta}\int_{0}^{1}dx\,\operatorname{Im}\left[\frac{f_s(2kx)-f_s(0)}{x}\left(1+\frac{1}{x}\right)^{2i\eta}{}_2F_1\left(i\eta,1+i\eta,1,\frac{1}{x^2}-i0\right)+\frac{f_s\left(\frac{2k}{x}\right)-f_s(0)}{x}(1+x)^{2i\eta}{}_2F_1\left(i\eta,1+i\eta,1,x^2-i0\right)\right]. \tag*{(51)}
A2s=−2η∫01dx sin⁡(ηlog⁡1+x1−x)x(x+1)[fs(2kx)−fs(2k)1−x−x2eπηfs(2kx)−fs(2k)1−x].(52)A_{2s}=-\frac{2}{\eta}\int_{0}^{1}dx\,\frac{\sin\left(\eta\log\frac{1+x}{1-x}\right)}{x(x+1)}\left[\frac{f_s\left(\frac{2k}{x}\right)-f_s(2k)}{1-x}-x^2e^{\pi\eta}\frac{f_s(2kx)-f_s(2k)}{1-x}\right]. \tag*{(52)}

The integrals can then be evaluated using the rectangle method or the trapezoidal rule, but since the individual function evaluations have a high cost in terms of CPU time and in practical applications such as optimization (fit) procedures, the final result has to be calculated several thousands of times at a given QQ, it is beneficial to search for methods requiring fewer evaluations. We find that the Gauss-Kronrod quadrature formula [33] provides an acceptable solution for this. This method works well in our case because the function to be integrated is changing fast near 0 and 1 but very slowly in the middle. Hence a varying bin width has to be applied, and such naturally arises from the Gauss-Kronrod iteration.

We utilized the boost C++ library to perform this integral and find that 15 nodes provide a fast converging result. The number of function evaluations and the integral results are shown in Figure 2. Here we chose a large QQ value, as the accuracy of the original numerical calculations, when one directly calculates the integral with ∣ψk(r)∣2|\psi_k(r)|^2, decreases towards large QQ, and hence accuracy can be well tested. The integral result changes on the order of 10−610^{-6}, and even on that scale, it converges when the number of maximum iterations is set to 3–4. It is also important to note that the typical order of magnitude of the bin-by-bin statistical uncertainties of a correlation function measured in high-energy experiments is significantly higher than the change in the integral values shown in Figure 2. If the tolerance parameter of the quadrature is larger than 10−310^{-3}, then the integral result differs slightly more from the plotted results. However, beyond that, if tolerance is at least 10−310^{-3}, increasing it further does not modify the integral result. On the other hand, the number of function evaluations increases fast, especially if the tolerance is very small and the number of iterations is large. Hence an optimal solution (analyzing results similar to Figure 2 for many RR, α\alpha, and QQ values) is provided by a tolerance value of 10−310^{-3} and a limit of 3 for the maximal number of iterations. With this, one integral requires up to a few hundred function evaluations, suitable for simultaneously fitting many experimental C2(Q)C_2(Q) datasets.

Figure 2. The result of the integral for R=6R=6 fm and α=1.4\alpha=1.4, at Q=0.2Q=0.2 GeV/cc (left) and the number of required function calls (right). Note that the integral result is represented as (1−C2(Q))⋅104(1-C_2(Q))\cdot10^4, so the differences are on the order of 10−610^{-6}. Both plots are shown as a function of the tolerance and the number of maximal iterations allowed in the Gauss-Kronrod integral.

Comparison with the original numerical integration method

After carefully testing the reliability of the Gauss-Kronrod numerical integral, we proceeded with comparing the new calculation to a previously utilized integral method based on Eq. (37), and used, e.g., in [18]. The previous method requires significantly more computational effort to achieve a reasonable precision at small α\alpha or large RR values. In that case, the values of the correlation function were pre-calculated for various α\alpha and RR values and saved in a large lookup table; from this table, an interpolation can be used to get the correlation function for any parameter combinations. In the following, let us call our new approach “wave function Fourier method” and denote the result with WFFC2(Q)\mathrm{WFF}C_2(Q). The previous method with the pre-calculated lookup table is denoted with tableC2(Q){}^{\mathrm{table}}C_2(Q).

To illustrate the difference between the two methods, we plot ΔC2(Q)=tableC2(Q)−WFFC2(Q)\Delta C_2(Q)={}^{\mathrm{table}}C_2(Q)-\mathrm{WFF}C_2(Q) for various α\alpha and RR values, as shown in Figure 3. We find that ΔC2(Q)\Delta C_2(Q) has a small dependence on QQ, generally its magnitude is larger at larger QQ values. It is also clear that the difference between the two methods is largest at the smallest α\alpha values, as expected. To better illustrate the dependence of ΔC2(Q)\Delta C_2(Q) on RR and α\alpha, in Figure 4 we plot the QQ-average ⟨ΔC2(Q)⟩\langle\Delta C_2(Q)\rangle values. It can be clearly seen that the ΔC2(Q)\Delta C_2(Q) difference between the two methods is smallest at large α\alpha and small RR values.

Figure 3. Difference between the correlation function calculated with a numerical integral method described in [18] (tableC2(Q){}^{\mathrm{table}}C_2(Q)) and the correlation function calculated with the wave-function Fourier method described in the current paper (WFFC2(Q){}^{\mathrm{WFF}}C_2(Q)). ΔC2(Q)\Delta C_2(Q) is plotted for 6 different α\alpha values and two RR values, for pions (left) and kaons (right) separately. At a given α\alpha value, ΔC2(Q)\Delta C_2(Q) goes smoothly through the shaded region when increasing RR values from R=3R=3 fm to R=12R=12 fm.

Figure 4. Relative momentum averaged ⟨ΔC2(Q)⟩\langle\Delta C_2(Q) \rangle as a function of RR and α\alpha for pions (left) and kaons (right).

The conclusion of this comparison is that, on one hand, we can be reasonably assured that our new methodology works well; as cautious users, we really strive for numerical verification, even if one is convinced by the

flawlessness of the mathematical derivation. On the other hand, our new methodology offers, at a significantly lower numerical cost, a way of calculating the C2(Q)C_2(Q) correlation function more robust than the previous method, especially at large QQ values. For this reason, we find that the new method is ready to be used in experimental analyses of measured C2(Q)C_2(Q) correlation functions.

Summary

In summary, we presented a novel method for calculating Bose-Einstein correlation functions with the Coulomb effect incorporated. Our method can be applied to any particle emitting source function assumptions that are expressed as a Fourier transform. In this way, our method starts directly from the interaction-free correlation function, which is the Fourier transform of the source function. The integrals necessary for our calculational scheme previously have eluded exact calculations, often resulting in experimental analyses using a fixed source size when correcting for the Coulomb effect. With precise data, however, this is no longer a viable method; our approach allows for simple and more exact handling of the Coulomb final state interactions. We demonstrated that for Lévy-stable sources (including the Gaussian limiting case), the results are close to those obtained with numerical integrals previously. On the other hand, at extreme values of the Lévy source parameters (small Lévy exponent α\alpha, and large Lévy scale RR), the new method is more precise and reliable. Furthermore, the new method allows for a direct and fast fitting of correlation functions obtained in experiments. We also published a software package [19] that is easily applicable and ready to use in experimental analyses.

The natural next step is to extend the methodology to the case when the assumption of spherical symmetry is rid of. In that case, one would have a final formula that is a linear functional of the Fourier transform of the pair source function, D(r)D(r), denoted by f(q)f(q) in this paper. It turns out that the necessary integrals (similar to but slightly more complicated than Eqs. (43) and (44) in this paper) are also readily calculable, and then the necessary limit when the regularizing parameter is removed (denoted as λ→0\lambda\to0 in this paper) is also manageable. The resulting formulas are, however (owing to the inherently more complex nature of the non-spherically symmetric case) somewhat more complicated, thus their analysis and implementation go beyond the scope of this paper.

Acknowledgements

This research was supported by the NKFIH grants K-136138 and TKP2021-NKTA-64.

Assorted mathematical and calculational details

In the Appendices, we present the calculations that were left out of the body of the paper. We rely heavily on complex analysis as well as on some intricacies of (Lebesgue) integrability; the latter of which we give a brief summary here. (A more detailed discussion can be found in any of the standard textbooks on mathematical analysis, such as Ref. [34].)

Discussion of Eq. (39)

The focal point of our method of circumventing the need for a back-and-forth integral transformation of the Fourier transform of the source function (denoted by f(q)f(q) in Eq. (33) and from that on) is Eq. (39), the mathematical justification of whose steps is thus essential. The main ingredients are conditions of integrability, Lebesgue’s dominated convergence theorem (or simply the Lebesgue theorem), and Fubini’s theorem. A summary of these theorems from a practical point of view is:

  • Concerning Lebesgue integrability, the key is the modulus of a function. A (real-valued, or complex-valued, or any normed space valued) function FF is integrable if and only if ∣F∣|F| is integrable. Also, if there is a GG function for which ∣G(x)∣≥∣F(x)∣|G(x)| \ge|F(x)| for almost all xx, and GG is integrable, then FF is also integrable. (Here, we denoted the integration variable by xx; it can be of any type of real integration variable.) Conversely, if ∣G(x)∣≥∣F(x)∣|G(x)| \ge|F(x)|, and FF is not integrable, then neither is GG.

  • Let Fλ(x)F_\lambda(x) be integrable functions, with λ\lambda as a parameter (either a continuous one or a discrete index), and for almost all xx let the pointwise limiting function F(x):=lim⁡λFλ(x)F(x) := \lim_\lambda F_\lambda(x) exist (for any reasonable type of limit; for example, λ→0\lambda\to0, or λ≡n→∞\lambda\equiv n \to\infty). Lebesgue’s theorem states that if there is a (λ\lambda-independent) GG function for which for almost any xx, ∣G(x)∣≥∣Fλ(x)∣|G(x)| \ge|F_\lambda(x)|, for any λ\lambda, then the integrals of the FλF_\lambda functions (the ∫Fλ\int F_\lambda values) converge, the limiting function FF is integrable, and the limit of the integrals is equal to the integral of the limiting function, lim⁡λ∫Fλ=∫F\lim_\lambda\int F_\lambda= \int F, i.e., the limit and the integral are interchangeable.

The key is the existence of the “dominant” GG function: in the case when there is no such function, none of the statements of the theorem are necessarily true. A simple (counter)example is the approximation of the Dirac delta with functions whose pointwise limit is everywhere zero; in this well-known case, one maybe does not even recognize this peculiarity. However, for less explored cases (like ours in this paper), one has to be careful when interchanging limits and integrals, and then this theorem comes in handy in many cases.

  • Fubini’s theorem concerns multiple integrals, iterated integrals, and the justification of interchange of integrals. The main point is if an F(x,y)F(x,y) function is such that its modulus is integrable in one order as a repeated integral; i.e., if the ∫dx(∫dy ∣F(x,y)∣)\int\mathrm{d}x \left(\int\mathrm{d}y\,|F(x,y)|\right) integral exists, then FF itself is integrable in both orders, and these integrals coincide: ∫dx∫dy F(x,y)=∫dy∫dx F(x,y)\int\mathrm{d}x \int\mathrm{d}y\,F(x,y) = \int\mathrm{d}y \int\mathrm{d}x\,F(x,y), i.e., the integrals are interchangeable.

The transformation in Eq. (39) is again written up, with the intermediate steps slightly more detailed, as

C2(Q)=∫d3r ∣ψk(r)∣2D(r)=1∫d3r lim⁡λ→0e−λr∣ψk(r)∣2D(r)=2lim⁡λ→0∫d3r e−λr∣ψk(r)∣2D(r)=3=3lim⁡λ→0∫d3r∫d3q(2π)3f(q)e−λr∣ψk(r)∣2eiqr=4lim⁡λ→0∫d3q(2π)3f(q)∫d3r e−λr∣ψk(r)∣2eiqr.(53)\begin{aligned} C_2(Q) &= \int\mathrm{d}^3r\,|\psi_k(r)|^2D(r) \overset{1}{=} \int\mathrm{d}^3r\,\lim_{\lambda\to0}e^{-\lambda r}|\psi_k(r)|^2D(r) \overset{2}{=} \lim_{\lambda\to0}\int\mathrm{d}^3r\,e^{-\lambda r}|\psi_k(r)|^2D(r) \overset{3}{=}\\ &\overset{3}{=} \lim_{\lambda\to0}\int\mathrm{d}^3r\int\frac{\mathrm{d}^3q}{(2\pi)^3}f(q)e^{-\lambda r}|\psi_k(r)|^2e^{i\mathbf{q}r} \overset{4}{=} \lim_{\lambda\to0}\int\frac{\mathrm{d}^3q}{(2\pi)^3}f(q)\int\mathrm{d}^3r\,e^{-\lambda r}|\psi_k(r)|^2e^{i\mathbf{q}r}. \tag*{(53)} \end{aligned}

In Step 1 we inserted the e−λre^{-\lambda r} regularizing factor; its limit being lim⁡λ→0e−λr=1\lim_{\lambda\to0}e^{-\lambda r}=1. Because D(r)D(r) is integrable, so is it multiplied by ∣ψk(r)∣2|\psi_k(r)|^2 (which is a bounded function), and thus ∣ψk(r)∣2⋅∣D(r)∣|\psi_k(r)|^2\cdot|D(r)| is a good dominant function (that is, integrable and greater or equal than the integrand, independent of λ\lambda). So by virtue of Lebesgue’s theorem, we could interchange the λ→0\lambda\to0 limit and the rr-integral in Step 2. In Step 3, we inserted the Fourier integral expression of D(r)D(r), careful (yet) about the order of integrals. However, since ∣eiqr∣=1|e^{i\mathbf{q}r}|=1, the double integral here has the modulus that is a product of ∣f(q)∣|f(q)|, an integrable qq-function (as assumed), and an integrable rr-function, e−λr∣ψk(r)∣2e^{-\lambda r}|\psi_k(r)|^2. So the repeated integral of the modulus exists, from which it follows (by Fubini’s theorem) that we can interchange the original integrals as well; this is Step 4 here.

In the resulting right-hand side, we cannot perform the λ→0\lambda\to0 limit in the integrands; that would mean the exchange of the original rr- and qq-integrals, which is not possible, as stated around Eq. (38). Instead, we have to calculate the integrals here, and then perform the λ→0\lambda\to0 limit.

Calculation of D1λs\mathcal{D}_{1\lambda s} and D2λs\mathcal{D}_{2\lambda s}

In this Appendix, we derive the Eqs. (45) and (46), the results of the integrals D1λs\mathcal{D}_{1\lambda s} and D2λs\mathcal{D}_{2\lambda s}, defined in Eqs. (43) and (44). We utilize Nordsieck’s method [31], who applied a similar technique to simplify a very similar integral that occurs in the theory of bremsstrahlung and pair creation. We write up the definitions (43)–(44) again, with a slight change of notation, to express the sin⁡(qr)\sin(qr) function with exponentials:

D1λs(q)=12i[D1λs(+)(q)−D1λs(−)(q)],D2λs(q)=12i[D2λs(+)(q)−D2λs(−)(q)],(54)\mathcal{D}_{1\lambda s}(q)=\frac{1}{2i}\left[\mathcal{D}_{1\lambda s}^{(+)}(q)-\mathcal{D}_{1\lambda s}^{(-)}(q)\right],\qquad\mathcal{D}_{2\lambda s}(q)=\frac{1}{2i}\left[\mathcal{D}_{2\lambda s}^{(+)}(q)-\mathcal{D}_{2\lambda s}^{(-)}(q)\right], \tag*{(54)}

where

D1λs(±)(q)=1q∫d3rre−λr±iqrM(1+iη,1,−i(kr+kr))M(1−iη,1,i(kr+kr)),(55)\mathcal{D}_{1\lambda_s}^{(\pm)}(q)=\frac{1}{q}\int\frac{\mathrm{d}^3r}{r}e^{-\lambda r\pm i\mathbf{q}r}M(1+i\eta,1,-i(kr+\mathbf{k}r))M(1-i\eta,1,i(kr+\mathbf{k}r)), \tag*{(55)}

and

D2λs(±)(q)=1q∫d3rre−λr±iqrM(1+iη,1,−i(kr−kr))M(1−iη,1,i(kr+kr)).(56)\mathcal{D}_{2\lambda_s}^{(\pm)}(q)=\frac{1}{q}\int\frac{\mathrm{d}^3r}{r}e^{-\lambda r\pm i\mathbf{q}r}M(1+i\eta,1,-i(kr-\mathbf{k}r))M(1-i\eta,1,i(kr+\mathbf{k}r)). \tag*{(56)}

We use the following complex contour integral representation of the confluent hypergeometric function:

M(a,1,z)=12πi∮(0+,1+)dttetz1t(1−1t)−a,(57)M(a,1,z)=\frac{1}{2\pi i}\oint^{(0+,1+)}\frac{\mathrm{d}t}{t}e^{tz}\frac{1}{t}\left(1-\frac{1}{t}\right)^{-a}, \tag*{(57)}

where the path is any closed curve that encircles the real line segment [0,1][0,1] once counterclockwise on the complex tt plane; this segment (a branch cut) is the only set on the tt plane where the integrand is not analytic.

Inserting twice this expression (with the integration variables denoted by tt and uu, whose paths are such as just specified) into Eqs. (55) and (56) for the two confluent hypergeometric functions in each, we get

D1λs(±)(q)=−14π2q∫d3rre−λr±iqr∮duu∮dtt(1−1t)−1−iη(1−1u)−1+iηe−i(t−u)(kr+kr),(58)\mathcal{D}_{1\lambda_s}^{(\pm)}(q)=-\frac{1}{4\pi^2q}\int\frac{\mathrm{d}^3r}{r}e^{-\lambda r\pm i\mathbf{q}r}\oint\frac{\mathrm{d}u}{u}\oint\frac{\mathrm{d}t}{t}\left(1-\frac{1}{t}\right)^{-1-i\eta}\left(1-\frac{1}{u}\right)^{-1+i\eta}e^{-i(t-u)(kr+\mathbf{k}r)}, \tag*{(58)}
D2λs(±)(q)=−14π2q∫d3rre−λr±iqr∮duu∮dtt(1−1t)−1−iη(1−1u)−1+iηe−i(t−u)kr+i(t+u)kr.(59)\mathcal{D}_{2\lambda_s}^{(\pm)}(q)=-\frac{1}{4\pi^2q}\int\frac{\mathrm{d}^3r}{r}e^{-\lambda r\pm i\mathbf{q}r}\oint\frac{\mathrm{d}u}{u}\oint\frac{\mathrm{d}t}{t}\left(1-\frac{1}{t}\right)^{-1-i\eta}\left(1-\frac{1}{u}\right)^{-1+i\eta}e^{-i(t-u)kr+i(t+u)\mathbf{k}r}. \tag*{(59)}

We would like to exchange the order of the integral over r\mathbf{r} and the contour integrals because for the r\mathbf{r}-integral, we could then use the following auxiliary formula, valid for any β∈C\beta\in\mathbb{C} complex number and B=(Bx,By,Bz)\mathbf{B}=(B_x,B_y,B_z) three-vector with any complex components:

∫d3rre−βr+Br=4πβ2−B2,if Re⁡β>∣Re⁡B∣.(60)\int\frac{\mathrm{d}^3r}{r}e^{-\beta r+\mathbf{B}r}=\frac{4\pi}{\beta^2-B^2},\qquad\text{if }\operatorname{Re}\beta>|\operatorname{Re}\mathbf{B}|. \tag*{(60)}

Here B2=Bx2+By2+Bz2B^2=B_x^2+B_y^2+B_z^2 even for complex BxB_x, ByB_y, and BzB_z (in particular, without complex conjugation), and the real part of the B\mathbf{B} vector is Re⁡B:=(Re⁡Bx,Re⁡By,Re⁡Bz)\operatorname{Re}\mathbf{B}:=(\operatorname{Re}B_x,\operatorname{Re}B_y,\operatorname{Re}B_z), thus ∣Re⁡B∣=(Re⁡Bx)2+(Re⁡By)2+(Re⁡Bz)2|\operatorname{Re}\mathbf{B}|=\sqrt{(\operatorname{Re}B_x)^2+(\operatorname{Re}B_y)^2+(\operatorname{Re}B_z)^2}. The condition written up in Eq. (60) is necessary and sufficient for the integral to exist. This is because a function is (Lebesgue) integrable if and only if its modulus is integrable, and

∣1re−βr+Br∣=1re−Re⁡βr+(Re⁡B)r,\left|\frac{1}{r}e^{-\beta r+\mathbf{B}r}\right|=\frac{1}{r}e^{-\operatorname{Re}\beta r+(\operatorname{Re}\mathbf{B})r},

which is integrable if and only if the multiplier of rr in the exponent, Re⁡β\operatorname{Re}\beta, is strictly bigger than the length of the vector there, ∣Re⁡B∣|\operatorname{Re}\mathbf{B}|. For real β\beta, BxB_x, ByB_y, BzB_z, the stated result is elementary, and because both the integral itself and the stated result are analytic functions of β\beta, BxB_x, ByB_y, BzB_z (provided the integral exists), the result is valid for all the allowed complex values, as stated.

If we can exchange the r\mathbf{r}-integral and the contour integrals in Eqs. (58) and (59), the we would get

D1λs(±)(q)?=−14π2q∮duu∮dtt(1−1t)−1−iη(1−1u)−1+iη∫d3rre−λr±iqr−i(t−u)(kr+kr),(61)\mathcal{D}_{1\lambda_s}^{(\pm)}(q)\mathrel{?}=-\frac{1}{4\pi^2q}\oint\frac{\mathrm{d}u}{u}\oint\frac{\mathrm{d}t}{t}\left(1-\frac{1}{t}\right)^{-1-i\eta}\left(1-\frac{1}{u}\right)^{-1+i\eta}\int\frac{\mathrm{d}^3r}{r}e^{-\lambda r\pm i\mathbf{q}r-i(t-u)(kr+\mathbf{k}r)}, \tag*{(61)}
D2λs(±)(q)?=−14π2q∮duu∮dtt(1−1t)−1−iη(1−1u)−1+iη∫d3rre−λr±iqr−i(t−u)kr+i(t+u)kr.(62)\mathcal{D}_{2\lambda_s}^{(\pm)}(q)\mathrel{?}=-\frac{1}{4\pi^2q}\oint\frac{\mathrm{d}u}{u}\oint\frac{\mathrm{d}t}{t}\left(1-\frac{1}{t}\right)^{-1-i\eta}\left(1-\frac{1}{u}\right)^{-1+i\eta}\int\frac{\mathrm{d}^3r}{r}e^{-\lambda r\pm i\mathbf{q}r-i(t-u)kr+i(t+u)\mathbf{k}r}. \tag*{(62)}

According to Fubini’s theorem, the interchange is justified if the modulus of the integrand is integrable in either order (provided that one has written up the contour integrals in a parametrized way, with integrals taken over line segment of the parameter). In our case, the moduli of the complex powers of tt and uu do not cause concern: since uu and tt run at a positive distance from the branch cut, these factors are bounded and thus do not spoil integrability. The factor of the modulus of the integrand that indeed has a role is that of the r\mathbf{r}-dependent part. In the case of D1λs\mathcal{D}_{1\lambda_s}, Eq. (61), the modulus of the part of the integrand that is of interest is

∣e−λr±iqr−i(t−u)(kr+kr)∣=e−[λ−kIm⁡(t−u)]r+Im⁡(t−u)kr,\left|e^{-\lambda r\pm i\mathbf{q}r-i(t-u)(kr+\mathbf{k}r)}\right|=e^{-[\lambda-k\operatorname{Im}(t-u)]r+\operatorname{Im}(t-u)\mathbf{k}r},

which is integrable over rr if and only if the following holds (by virtue of the condition stated in Eq. (60)):

λk−Im⁡(t−u)>∣Im⁡(t−u)∣⟺Im⁡(t−u)<λ2k(for D1λs).(63)\frac{\lambda}{k}-\operatorname{Im}(t-u)>|\operatorname{Im}(t-u)|\qquad\Longleftrightarrow\qquad\operatorname{Im}(t-u)<\frac{\lambda}{2k}\qquad\text{(for }\mathcal{D}_{1\lambda_s}\text{)}. \tag*{(63)}

We arrived at the simpler condition by sorting through all the different possibilities for Im⁡u\operatorname{Im}u and Im⁡t\operatorname{Im}t. Similarly, in the case of D2λs\mathcal{D}_{2\lambda_s}, Eq. (62), we have

∣e−λr±iqr−i(t−u)kr+i(t+u)kr∣=e−[λ−kIm⁡(t−u)]r−Im⁡(t+u)kr,\left|e^{-\lambda r\pm i\mathbf{q}r-i(t-u)kr+i(t+u)\mathbf{k}r}\right|=e^{-[\lambda-k\operatorname{Im}(t-u)]r-\operatorname{Im}(t+u)\mathbf{k}r},

which is integrable over rr if and only if

λk−Im⁡(t−u)>∣Im⁡(t+u)∣⟺Im⁡u>−λ2kandIm⁡t<λ2k(64)\frac{\lambda}{k}-\operatorname{Im}(t-u)>\left|\operatorname{Im}(t+u)\right| \quad\Longleftrightarrow\quad\operatorname{Im}u>-\frac{\lambda}{2k}\quad\text{and}\quad\operatorname{Im}t<\frac{\lambda}{2k} \tag*{(64)}

(for D2λs\mathcal{D}_{2\lambda s}).

If these conditions hold for all possible uu and tt, then the results of these rr-integrals are continuous bounded functions of uu and tt, so their contour integrals exist. So the condition for the integrals over rr and u,tu,t to be interchangeable is that the inequalities stated in Eqs. (63) and (64) hold for any uu and tt on their integration paths. Thus if we require these additional constraints on the uu- and tt-paths, we can interchange the integrals, and utilizing Eq. (60), we get the following from Eqs. (61) and (62):

D1λs(±)(q)=−1πq1λ∓iq∮dtt(1−1t)−1−iη∮duu(1−1u)−1+iη1λ∓iq+2ik(t−u)(65)\mathcal{D}^{(\pm)}_{1\lambda s}(q)=-\frac{1}{\pi q}\frac{1}{\lambda\mp iq}\oint\frac{\mathrm{d}t}{t}\left(1-\frac{1}{t}\right)^{-1-i\eta}\oint\frac{\mathrm{d}u}{u}\left(1-\frac{1}{u}\right)^{-1+i\eta}\frac{1}{\lambda\mp iq+2ik(t-u)} \tag*{(65)}
D2λs(±)(q)=−1πq∮dtt(1−1t)−1−iη12kt−i(λ∓iq)∮duu(1−1u)−1+iη12ku+i(λ∓iq)(66)\mathcal{D}^{(\pm)}_{2\lambda s}(q)=-\frac{1}{\pi q}\oint\frac{\mathrm{d}t}{t}\left(1-\frac{1}{t}\right)^{-1-i\eta}\frac{1}{2kt-i(\lambda\mp iq)}\oint\frac{\mathrm{d}u}{u}\left(1-\frac{1}{u}\right)^{-1+i\eta}\frac{1}{2ku+i(\lambda\mp iq)} \tag*{(66)}

where the paths of the uu and tt variables run as specified: they both encircle the [0,1][0,1] branch cut and obey at all points the conditions (63) in case of D1λs\mathcal{D}_{1\lambda s}, and (64) in case of D2λs\mathcal{D}_{2\lambda s}, respectively. (For any finite λ\lambda and kk, it is indeed possible to specify the integration paths in this way.) The case of D2λs\mathcal{D}_{2\lambda s} is simpler because the integrand factorizes in uu and tt; we have written Eq. (66) in a way that highlights this.

The next step is to perform the uu-integral at a fixed tt. As a function of uu, the integrand in both cases is analytic everywhere except for the branch cut along u∈[0,1]u\in[0,1] which is encircled by the path, as well as for a simple pole, denoted in case of D1λs\mathcal{D}_{1\lambda s} and D2λs\mathcal{D}_{2\lambda s} by u1s(t)u_{1s}(t) and u2s(t)u_{2s}(t), respectively:

u1s(t)=t∓q2k−iλ2k,u2s(t)=−iλ2k−q2k(67)u_{1s}(t)=t\mp\frac{q}{2k}-\frac{i\lambda}{2k},\qquad u_{2s}(t)=-\frac{i\lambda}{2k}-\frac{q}{2k} \tag*{(67)}

We see that Im⁡(t−u1s)=λ2k\operatorname{Im}(t-u_{1s})=\frac{\lambda}{2k} and Im⁡(u2s)=−λ2k\operatorname{Im}(u_{2s})=-\frac{\lambda}{2k}: this means that (for any tt that is in its allowed domain) u1su_{1s} and u2su_{2s} do not satisfy the conditions in Eqs. (63) and (64). From this, one concludes that u1su_{1s} and u2su_{2s} lies outside of the integration contour on the uu plane: were it otherwise, a narrower integration path chosen for uu could cross u1su_{1s} and u2su_{2s}, but this is impossible because then u1su_{1s} and u2su_{2s} could not violate the conditions.

One can then expand the uu-integration contour to infinity: the integrand decreases rapidly enough (as ∼1u2\sim\frac{1}{u^2} in both cases), so the non-vanishing contribution comes from the simple poles u1su_{1s} and u2su_{2s}, so from the residue theorem we have (with an extra minus sign from the negative sense of the paths encircling the poles)

D1λs(±)(q)=−1q∮dtt(1−1t)−1−iηk(λ∓iq)1u1s(1−1u1s)−1+iη(68)\mathcal{D}^{(\pm)}_{1\lambda s}(q)=-\frac{1}{q}\oint\frac{\mathrm{d}t}{t}\frac{\left(1-\frac{1}{t}\right)^{-1-i\eta}}{k(\lambda\mp iq)}\frac{1}{u_{1s}}\left(1-\frac{1}{u_{1s}}\right)^{-1+i\eta} \tag*{(68)}
D2λs(±)(q)=−1q2λ∓iq(λ∓iq−2ikλ∓iq)−1+iη∮dtt(1−1t)−1−iη2kt−i(λ∓iq)(69)\mathcal{D}^{(\pm)}_{2\lambda s}(q)=-\frac{1}{q}\frac{2}{\lambda\mp iq}\left(\frac{\lambda\mp iq-2ik}{\lambda\mp iq}\right)^{-1+i\eta}\oint\frac{\mathrm{d}t}{t}\frac{\left(1-\frac{1}{t}\right)^{-1-i\eta}}{2kt-i(\lambda\mp iq)} \tag*{(69)}

where in case of D1λs\mathcal{D}_{1\lambda s}, u1s(t)u_{1s}(t) is as in Eq. (67), while in case of D2λs\mathcal{D}_{2\lambda s} we inserted u2su_{2s} from Eq. (67) right away.5

In Eq. (69), we can use this same method for the tt-integral. The integrand has a simple pole in tt at t2s=iλ2k+q2kt_{2s}=\frac{i\lambda}{2k}+\frac{q}{2k}, and the same reasoning as for u2su_{2s} above yields that because Im⁡t2s=λ2k\operatorname{Im}t_{2s}=\frac{\lambda}{2k}, t2st_{2s} does not satisfy the condition in Eq. (64), so it lies outside of the integration contour. Thus an expansion of the contour to infinity yields (again because of the ∼1t2\sim\frac{1}{t^2} decrease) just the contribution from the pole at t2st_{2s}. From this, we get

D2λs(±)(q)=4πq1λ∓iq(λ∓iq−2ikλ∓iq)−1+iη1λ∓iq(λ∓iq+2ikλ∓iq)−1−iη(70)\mathcal{D}^{(\pm)}_{2\lambda s}(q)=\frac{4\pi}{q}\frac{1}{\lambda\mp iq}\left(\frac{\lambda\mp iq-2ik}{\lambda\mp iq}\right)^{-1+i\eta}\frac{1}{\lambda\mp iq}\left(\frac{\lambda\mp iq+2ik}{\lambda\mp iq}\right)^{-1-i\eta} \tag*{(70)}

This leads to the result for D2λs\mathcal{D}_{2\lambda s}, stated in Eq. (46) in the main text, after some simplification.

In the case of D1λs\mathcal{D}_{1\lambda s}, the remaining integrand in Eq. (68) does not have a simple pole as a function of tt (besides the branch cut on [0,1][0,1]), but another branch cut on a straight line segment between t=±q2k+iλ2kt=\pm\frac{q}{2k}+\frac{i\lambda}{2k} and t=1±q2k+iλ2kt=1\pm\frac{q}{2k}+\frac{i\lambda}{2k} (i.e., where u1s(t)=0u_{1s}(t)=0 and u1s(t)=1u_{1s}(t)=1). This branch cut, as its imaginary part, λ2k\frac{\lambda}{2k} is in violation of condition (63), lies outside the tt integration contour. In this case, we can transform the integral in Eq. (68) into a form that yields an ordinary hypergeometric function. To this end, we substitute t=1st=\frac{1}{s}, with which we arrive at

D1λs(±)(q)=N∮ds (1−s)−1−iηs−A[βs−Bs−A]−1+iη,(71)\mathcal{D}_{1\lambda s}^{(\pm)}(q)=N\oint ds\,\frac{(1-s)^{-1-i\eta}}{s-A}\left[\beta\frac{s-B}{s-A}\right]^{-1+i\eta}, \tag*{(71)}

where we used the notations

N=−2iq(λ∓iq)2,A:=2kiλ∓q,B:=2kiλ∓q+2k,β:=1+2kiλ∓q.(72)N=\frac{-2i}{q(\lambda\mp iq)^2},\qquad A:=\frac{2k}{i\lambda\mp q},\qquad B:=\frac{2k}{i\lambda\mp q+2k},\qquad\beta:=1+\frac{2k}{i\lambda\mp q}. \tag*{(72)}

and the integration contour on the ss plane does not intersect the [1,∞][1,\infty] half line (the image of the original t∈[0,1]t\in[0,1] cut), but encircles s=0s=0 (the image of t=∞t=\infty) and the cut between s=As=A and s=Bs=B (which is a circle segment; the image of the outlying branch cut on the tt plane).

We need now a linear transformation of the integration variable s→τs\to\tau, so that the three singular points s=1s=1, s=As=A and s=Bs=B transform into τ=0\tau=0, τ=1\tau=1 and τ=x\tau=x with some xx. This xx will then be the variable of the resulting hypergeometric function; this is motivated by the integral representation

2F1(a,b,c,z)=Γ(c)Γ(b−c+1)Γ(b)12πi∫τ=−∞c(0+,z+)dτ τa−c(1−τ)c−b−1(τ−x)−a,(73){}_2F_1(a,b,c,z)=\frac{\Gamma(c)\Gamma(b-c+1)}{\Gamma(b)}\frac{1}{2\pi i}\int_{\tau=-\infty}^{c(0+,z+)}d\tau\,\tau^{a-c}(1-\tau)^{c-b-1}(\tau-x)^{-a}, \tag*{(73)}

with the τ\tau integration path coming from Re⁡τ=−∞\operatorname{Re}\tau=-\infty, going back there, while encircling the half line branch cuts τ∈R0−\tau\in\mathbb{R}_0^{-} and τ−x∈R0−\tau-x\in\mathbb{R}_0^{-}, but not intersecting τ∈[1,∞]\tau\in[1,\infty], the third branch cut of the integrand.6

The most convenient s→τs\to\tau transformation (among the many possible ones) turns out to be the following:

s=B+(1−B)τ;(74)s=B+(1-B)\tau; \tag*{(74)}

this takes s=Bs=B into τ=0\tau=0, s=1s=1 into τ=1\tau=1, and s=As=A into τ=A−B1−B\tau=\frac{A-B}{1-B}, and we get

D1λs=N∮(0+,x+)dτ ((1−B)(1−τ))−1−iη1τ−x(βττ−x)−1+iη,withx=A−B1−B=4k2(q±iλ)2.(75)\mathcal{D}_{1\lambda s}=N\oint^{(0+,x+)}d\tau\,((1-B)(1-\tau))^{-1-i\eta}\frac{1}{\tau-x}\left(\frac{\beta\tau}{\tau-x}\right)^{-1+i\eta}, \qquad\text{with}\quad x=\frac{A-B}{1-B}=\frac{4k^2}{(q\pm i\lambda)^2}. \tag*{(75)}

Now in order to arrive at a form resembling the integral representation (73), we have to simplify the complex powers. This is not entirely trivial because owing to the branch cut of power functions, for complex numbers, the identity (ab)c=acbc(ab)^c=a^cb^c holds only if arg⁡a+arg⁡b=arg⁡(ab)\arg a+\arg b=\arg(ab), i.e., if −π<arg⁡a+arg⁡b≤π-\pi<\arg a+\arg b\leq\pi. If the displayed quantities are functions of some other complex variable tt, it is very cumbersome to verify this condition directly. For line integrals, a workaround is that if the identity (ab)c=acbc(ab)^c=a^cb^c is verified in a small neighborhood of some point tt on the path, and both sides are analytic in tt in the vicinity of the path (practically: the path does not intersect branch cuts of either form on the tt plane), then owing to the rigidity of analytic functions, the substitution is justified along the whole integration path.

In the case of Eq. (75), the steps are as follows. We do not write up all the details about these transformations here; the interested reader is advised to directly check that these steps are indeed justified.

  1. First we do the ((1−B)(1−τ))−1−iη=(1−B)−1−iη(1−τ)−1−iη((1-B)(1-\tau))^{-1-i\eta}=(1-B)^{-1-i\eta}(1-\tau)^{-1-i\eta} substitution. For this, we must verify that the integration path on the τ\tau plane can be specified (and we indeed do so) in a way that it does not intersect the branch cuts of either of these forms (half lines, but in different directions).

  1. Next, we substitute (βττ−x)−1+iη\left(\frac{\beta\tau}{\tau-x}\right)^{-1+i\eta} with β−1+iη(ττ−x)−1+iη\beta^{-1+i\eta}\left(\frac{\tau}{\tau-x}\right)^{-1+i\eta}. For this, we must verify that we can take the path so that it does not intersect the branch cuts of either form (a circle segment and a line segment, respectively); for this, the key element is that none of these cuts intersect with the other branch cuts of the integrand.

  1. Now we open the contour so that instead of being a closed one encircling 00 and xx, it comes from Re⁡τ=−∞\operatorname{Re}\tau=-\infty and returns there. This can be done because, owing to the exponents, the integrand decreases fast enough (as ∼1τ2\sim\frac{1}{\tau^2}) at ∣τ∣→∞|\tau|\to\infty.

  1. Finally, we can write (ττ−x)−1+iη=τ−1+iη(τ−x)1−iη\left(\frac{\tau}{\tau-x}\right)^{-1+i\eta}=\tau^{-1+i\eta}(\tau-x)^{1-i\eta}; besides τ∈[1,∞]\tau\in[1,\infty], this latter form has branch cuts that are straight half lines from Re⁡τ=−∞\operatorname{Re}\tau=-\infty to 00 and xx, respectively.

Finally, we thus arrive at the following expression, where we can utilize the (73) integral representation:

D1λs=Nβ−1+iη(1−B)1+iη∫τ=−∞(0+,x+)dτ τ−1+iη(1−τ)−1−iη(τ−x)−iη=2πiNβ−1+iη(1−B)1+iη 2F1(iη,1+iη,1,x),(76)D_{1\lambda s}=N\frac{\beta^{-1+i\eta}}{(1-B)^{1+i\eta}}\int_{\tau=-\infty}^{(0+,x+)}d\tau\,\tau^{-1+i\eta}(1-\tau)^{-1-i\eta}(\tau-x)^{-i\eta}=2\pi iN\frac{\beta^{-1+i\eta}}{(1-B)^{1+i\eta}}\,{}_2F_1(i\eta,1+i\eta,1,x), \tag*{(76)}

and from this we get the result given for D1λsD_{1\lambda s} in the body of the paper, Eq. (45), by substituting NN, β\beta, BB, and xx, and finally combining together D1λs(+)D_{1\lambda s}^{(+)} and D1λs(−)D_{1\lambda s}^{(-)} again, according to Eq. (55).

Derivation of the final formula, Eq. (48)

In this Appendix, we derive Eq. (48) along the expression of the quantities denoted by A1sA_{1s} and A2sA_{2s} in the body of the paper, Eqs. (49) and (50). The starting point is Eq. (42), the expression of C2(Q)C_2(Q) with D1λsD_{1\lambda s} and D2λsD_{2\lambda s} as a λ→0\lambda\to0 limit. We write this up again but first with the integral denoted as a three-dimensional qq-integral, and perform the first crucial step as we “separate” the fs(0)f_s(0) and fs(2k)f_s(2k) values. This idea comes from observing that the values of fs(q)f_s(q) at q=0q=0 and q=2kq=2k play a special role in the interaction-free case as C2(0)(Q)=fs(0)+fs(2k)C_2^{(0)}(Q)=f_s(0)+f_s(2k), moreover, fs(0)f_s(0) and fs(2k)f_s(2k) enter by virtue of the “plain” and the “cross” terms from the modulus square of the wave function, just as D1λsD_{1\lambda s} and D2λsD_{2\lambda s} in the interacting case. We thus write

C2(Q)=∣N∣28π3lim⁡λ→0∫d3q fs(q)[D1λs(q)+D2λs(q)]==∣N∣28π3fs(0)lim⁡λ→0∫d3q D1λs(q)+∣N∣28π3lim⁡λ→0∫d3q [fs(q)−fs(0)]D1λs(q)++∣N∣28π3fs(2k)lim⁡λ→0∫d3q D2λs(q)+∣N∣28π3lim⁡λ→0∫d3q [fs(q)−fs(2k)]D2λs(q).(77)\begin{aligned} C_2(Q)&=\frac{|\mathcal{N}|^2}{8\pi^3}\lim_{\lambda\to0}\int d^3q\,f_s(q)\left[D_{1\lambda s}(q)+D_{2\lambda s}(q)\right]=\\ &=\frac{|\mathcal{N}|^2}{8\pi^3}f_s(0)\lim_{\lambda\to0}\int d^3q\,D_{1\lambda s}(q)+\frac{|\mathcal{N}|^2}{8\pi^3}\lim_{\lambda\to0}\int d^3q\,[f_s(q)-f_s(0)]D_{1\lambda s}(q)+\\ &\quad+\frac{|\mathcal{N}|^2}{8\pi^3}f_s(2k)\lim_{\lambda\to0}\int d^3q\,D_{2\lambda s}(q)+\frac{|\mathcal{N}|^2}{8\pi^3}\lim_{\lambda\to0}\int d^3q\,[f_s(q)-f_s(2k)]D_{2\lambda s}(q). \tag*{(77)} \end{aligned}

Now for λ>0\lambda>0, the integrals of D1λsD_{1\lambda s} and D2λsD_{2\lambda s}, denoted temporarily by I1I_1 and I2I_2, do exist:

I1:=∫d3q D1λs(q)=4π∫0∞dq q2D1λs(q),I2:=∫d3q D2λs(q)=4π∫0∞dq q2D2λs(q),(78)I_1:=\int d^3q\,D_{1\lambda s}(q)=4\pi\int_0^\infty dq\,q^2D_{1\lambda s}(q),\qquad I_2:=\int d^3q\,D_{2\lambda s}(q)=4\pi\int_0^\infty dq\,q^2D_{2\lambda s}(q), \tag*{(78)}

which can be directly seen from their expressions, Eqs. (45) and (46): q2D1λs(q)q^2D_{1\lambda s}(q) and q2D2λs(q)q^2D_{2\lambda s}(q) are continuous and bounded on every compact [0,qmax⁡][0,q_{\max}] interval, and they decrease as ∼1q2\sim\frac{1}{q^2}, owing to the taking of the imaginary part in them; recall that 2F1(a,b,c,x)=1+abcx+…{}_2F_1(a,b,c,x)=1+\frac{ab}{c}x+\ldots as x→0x\to0. So the I1I_1 and I2I_2 integrals are given by the value of the Fourier transforms of D1λsD_{1\lambda s} and D2λsD_{2\lambda s} at r=0r=0 (where rr is the variable of their Fourier transforms).7 But we actually calculated D1λsD_{1\lambda s} and D2λsD_{2\lambda s} as Fourier transforms, so we just need to evaluate the integrands of the defining (43)–(44) integrals at r=0r=0, from which, knowing that M(a,b,x=0)=1M(a,b,x=0)=1, we have I1=I2=8π3I_1=I_2=8\pi^3, so

C2(Q)=∣N∣2[fs(0)+fs(2k)+lim⁡λ→0∫0∞dq q2D1λs(q)2π2[fs(q)−fs(0)]+lim⁡λ→0∫0∞dq q2D2λs(q)2π2[fs(q)−fs(2k)]].(79)C_2(Q)=|\mathcal{N}|^2\left[f_s(0)+f_s(2k)+\lim_{\lambda\to0}\int_0^\infty dq\,\frac{q^2D_{1\lambda s}(q)}{2\pi^2}[f_s(q)-f_s(0)]+\lim_{\lambda\to0}\int_0^\infty dq\,\frac{q^2D_{2\lambda s}(q)}{2\pi^2}[f_s(q)-f_s(2k)]\right]. \tag*{(79)}

Our statement is now that these remaining limits result in A1sA_{1s} and A2sA_{2s} as written up in Eqs. (49) and (50). Substituting D1λsD_{1\lambda s} and D2λsD_{2\lambda s} from Eqs. (45) and (46), and using the convenient F+(x)≡2F1(iη,1+iη,1,x)F_+(x)\equiv{}_2F_1(i\eta,1+i\eta,1,x) notation, we thus have to prove that

lim⁡λ→0∫0∞dq [fs(q)−fs(0)]Im⁡[(1+2kq+iλ)2iη(λ−iq)2F+(4k2(q+iλ)2)]=∫0∞dq fs(q)−fs(0)qIm⁡[(1+2kq)2iηF+(4k2q2−i0)],(80)\lim_{\lambda\to0}\int_0^\infty dq\,[f_s(q)-f_s(0)]\operatorname{Im}\left[\frac{\left(1+\frac{2k}{q+i\lambda}\right)^{2i\eta}}{(\lambda-iq)^2}F_+\left(\frac{4k^2}{(q+i\lambda)^2}\right)\right]=\int_0^\infty dq\,\frac{f_s(q)-f_s(0)}{q}\operatorname{Im}\left[\left(1+\frac{2k}{q}\right)^{2i\eta}F_+\left(\frac{4k^2}{q^2}-i0\right)\right], \tag*{(80)}
lim⁡λ→0∫0∞dq [fs(q)−fs(2k)]Im⁡[(λ−iq−2ik)iη(λ−iq+2ik)−iη(λ−iq)2+4k2]=−∫0∞dq fs(q)−fs(2k)q−2kqq+2kIm⁡[(q+2k)iη(q−2k+i0)iη].(81)\lim_{\lambda\to0}\int_0^\infty dq\,[f_s(q)-f_s(2k)]\operatorname{Im}\left[\frac{(\lambda-iq-2ik)^{i\eta}(\lambda-iq+2ik)^{-i\eta}}{(\lambda-iq)^2+4k^2}\right]=-\int_0^\infty dq\,\frac{f_s(q)-f_s(2k)}{q-2k}\frac{q}{q+2k}\operatorname{Im}\left[\frac{(q+2k)^{i\eta}}{(q-2k+i0)^{i\eta}}\right]. \tag*{(81)}

Note that the pointwise λ→0\lambda\to0 limit of the integrands on the left-hand sides are just the ones on the right-hand sides. We want to apply Lebesgue’s theorem (see Appendix § 7.§ 1 above) so that we can interchange the limit and the integral. To this end, we need a “dominant” function, i.e., an integrable function whose modulus is greater or equal to that of the integrand, independently of λ\lambda, for any λ>0\lambda>0. We do not explicitly write up this function, however, we give the various estimations that are needed to construct this function.

  1. The fsf_s function is bounded everywhere, fs(q)≤Kf_s(q) \le K. Also, the main point of introducing the subtraction of fs(0)f_s(0) and fs(2k)f_s(2k) values as in Eqs. (80) and (81) is that for the fractions appearing from these can be dominated by a function that is integrable around q=0q=0 and q=2kq=2k if fsf_s is continuously differentiable everywhere: in this case for a finite qmax⁡q_{\max} value,

∣f(q)−f(0)∣q,and∣f(q)−f(2k)∣q−2k are bounded on [0,qmax⁡],(82)\frac{\lvert f(q)-f(0)\rvert}{q}, \qquad\text{and} \qquad\frac{\lvert f(q)-f(2k)\rvert}{q-2k}\ \text{are bounded on }[0,q_{\max}], \tag*{(82)}

and thus integrable here. Had we not have subtracted fs(0)f_s(0) and fs(2k)f_s(2k), the fractions fs(q)q\frac{f_s(q)}{q} and fs(q)q−2k\frac{f_s(q)}{q-2k} would have not be integrable around q=0q=0 and q=2kq=2k.

In the case of our preeminent practical example of source functions, the Lévy distribution, fs(q)f_s(q) is not necessarily differentiable around q=0q=0. However, for our estimation to find the dominant function here, it is enough if the fractions written up in Eq. (82) are not bounded but can be dominated by an integrable function. This is readily satisfied also in the case of Lévy distributions, where the following estimation is true instead:

For Lévy distributions, for any qq and q′q', ∣f(q)−f(q′)∣≤K′∣q−q′∣α\lvert f(q)-f(q')\rvert\le K'\lvert q-q'\rvert^\alpha, this is enough. {#eq:83}

  1. For the various factors appearing, the following estimations can be made. For real XX and YY, the pure imaginary power XiYX^{iY} is bounded by e−πYe^{-\pi Y} from below and by eπYe^{\pi Y} from above. The F+(x)\mathcal{F}_+(x) hypergeometric function is bounded on the whole complex xx plane.8

  1. Owing to the continuous differentiability of F+(x)\mathcal{F}_+(x) around x=0x=0, there exists a C′C' constant for which ∣F+(x)−1∣≤C′∣x∣\lvert\mathcal{F}_+(x)-1\rvert\le C'\lvert x\rvert for small enough xx. Also, for pure imaginary powers, the following estimation can be made: if Y∈RY \in\mathbb{R}, 0≤R<10 \le R < 1 and ∣z∣<R\lvert z\rvert< R, then ∣(1+z)iY−1∣≤11−Reπ∣Y∣∣Y∣⋅∣z∣\lvert(1+z)^{iY}-1\rvert\le\frac{1}{1-R}e^{\pi\lvert Y\rvert}\lvert Y\rvert\cdot\lvert z\rvert holds.9

From these, with some effort, one can construct the dominant function; this is best done by separately treating the q∈[0,Qmax⁡]q \in[0,Q_{\max}] and the q∈[Qmax⁡,∞]q \in[Q_{\max},\infty] intervals, where Qmax⁡Q_{\max} is an arbitrary value that is at a safe distance above 2k2k; say: Qmax⁡=4kQ_{\max}=4k. In this way, the exchange of the integral and the limit in Eqs. (80) and (81) becomes justified, which leads to the desired results for A1s\mathcal{A}_{1s} and A2s\mathcal{A}_{2s}.

References

  1. [1]J. Adams et al. (STAR), Nucl. Phys. A 757, 102 (2005), arXiv:nucl-ex/0501009DOI
  2. [2]K. Adcox et al. (PHENIX Coll.), Nucl.Phys. A 757, 184 (2005), arXiv:nucl-ex/0410003DOI
  3. [3]B.B. Back et al. (PHOBOS), Nucl. Phys. A 757, 28 (2005), arXiv:nucl-ex/0410022
  4. [4]I. Arsene et al. (BRAHMS), Nucl. Phys. A 757, 1 (2005), arXiv:nucl-ex/0410020DOI
  5. [5]R. Hanbury Brown, R.Q. Twiss, Nature 178, 1046 (1956)
  6. [6]M.A. Lisa, S. Pratt, R. Soltz, U. Wiedemann, Ann. Rev. Nucl. Part. Sci. 55, 357 (2005), arXiv:nucl-ex/0505014DOI
  7. [7]G. Goldhaber, S. Goldhaber, W.Y. Lee, A. Pais, Phys. Rev. 120, 300 (1960)
  8. [8]S.S. Adler et al. (PHENIX), Phys. Rev. Lett. 93, 152302 (2004), arXiv:nucl-ex/0401003DOI
  9. [9]J. Adams et al. (STAR), Phys. Rev. C 71, 044906 (2005), arXiv:nucl-ex/0411036
  10. [10]M.A. Lisa, U.W. Heinz, U.A. Wiedemann, Phys. Lett. B 489, 287 (2000), arXiv:nucl-th/0003022DOI
  11. [11]M. Csanád, B. Tomasik, T. Csörgő, Eur. Phys. J. A 37, 111 (2008), arXiv:0801.4434DOI
  12. [12]S. Pratt, Phys. Rev. Lett. 102, 232301 (2009), arXiv:0811.3363DOI
  13. [13]S.S. Adler et al. (PHENIX), Phys. Rev. Lett. 98, 132301 (2007), arXiv:nucl-ex/0605032DOI
  14. [14]A. Kisiel, D.A. Brown, Phys. Rev. C 80, 064911 (2009), arXiv:0901.3527DOI
  15. [15]A. Adare et al. (PHENIX), Phys. Rev. C 97, 064911 (2018), arXiv:1709.05649DOI
  16. [16]H. Adhikary et al. (NA61/SHINE) (2023), arXiv:2302.04593DOI
  17. [17]A. Tumasyan et al. (CMS) (2023), arXiv:2306.11574
  18. [18]D. Kincses, M.I. Nagy, M. Csanád, Phys. Rev. C 102, 064912 (2020), arXiv:1912.01381
  19. [19]Correlation function calculation with Lévy source and Coulomb FSI (2023), https://https://github.com/csanadm/CoulCorrLevyIntegral
  20. [20]F.B. Yano, S.E. Koonin, Phys. Lett. B 78, 556 (1978)
  21. [21]S. Pratt, Phys. Rev. C 56, 1095 (1997)
  22. [22]D.H. Rischke, M. Gyulassy, Nucl. Phys. A 608, 479 (1996), arXiv:nucl-th/9606039DOI
  23. [23]L.D. Landau, E.M. Lifshits, Quantum Mechanics: Non-Relativistic Theory, Vol. v.3 of Course of Theoretical Physics (Butterworth-Heinemann, Oxford, 1991), ISBN 978-0-7506-3539-4
  24. [24]T. Csörgő, S. Hegyi, W.A. Zajc, Eur. Phys. J. C 36, 67 (2004), arXiv:nucl-th/0310042DOI
  25. [25]T. Csörgő, S. Hegyi, T. Novak, W.A. Zajc, Acta Phys. Polon. B 36, 329 (2005), arXiv:hep-ph/0412243
  26. [26]T. Csörgő, S. Hegyi, T. Novak, W.A. Zajc, AIP Conf. Proc. 828, 525 (2006), arXiv:nucl-th/0512060DOI
  27. [27]J. Cimermāň, C. Plumberg, B. Tomášik, Phys. Part. Nucl. 51, 282 (2020), arXiv:1909.07998arxiv.org/abs/1909.07998
  28. [28]M. Csanád, T. Csörgő, M. Nagy, Braz. J. Phys. 37, 1002 (2007), arXiv:hep-ph/0702032DOI
  29. [29]D. Kincses, M. Stefaniak, M. Csanád, Entropy 24, 308 (2022), arXiv:2201.07962
  30. [30]M. Csanád, S. Lökös, M. Nagy, Universe 5, 133 (2019), arXiv:1905.09714DOI
  31. [31]A. Nordsieck, Phys. Rev. 93, 785 (1954)
  32. [32]DLMF, NIST Digital Library of Mathematical Functions, https://dlmf.nist.gov/, Release 1.1.10 of 2023-06-15, f. W. J. Oliver, A. B. Olde Daalhuis, D. W. Lozier, B. I. Schneider, R. F. Boisvert, C. W. Clark, B. R. Miller, B. V. Saunders, H. S. Cohl, and M. A. McClain, eds., https://dlmf.nist.gov/
  33. [33]P. Davis, P. Rabinowitz, W. Rheinbolt, Methods of Numerical Integration, Computer Science and Applied Mathematics (Elsevier Science, 2014), ISBN 9781483264288, https://books.google.com/books?id=mbLiBQAAQBAJ
  34. [34]W. Rudin, Real and Complex Analysis (McGraw-Hill Science/Engineering/Math, 1986), ISBN 0070542341, http://www.amazon.com/exec/obidos/redirect?tag=citeulike07-20&path=ASIN/0070542341

Paper details

vlnc:W_zu2n4chn
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