betaFeedback

WorksPhysics and Astronomy

Spiral momentum spectra and spin asymmetry in vacuum pair production driven by phase-modulated elliptically polarized electric fields

Abstract

We study vacuum electron–positron pair production in phase-modulated elliptically polarized electric fields using the quantum kinetic equations. We analyze how the modulation amplitude, frequency, ellipticity, and modulation form affect the spin-resolved momentum spectra and the spin asymmetry. Cosine-type modulation can deform ring-shaped momentum distributions into spiral patterns. This deformation depends nonmonotonically on the modulation amplitude and frequency and is most pronounced for circular polarization. In the transverse momentum plane, the spin-resolved yields favor the spin channel along the direction of field handedness. Reversing the field handedness would reverse the favored channel while leaving the total yield unchanged. Sine-type modulation preserves ring-shaped and mirror-symmetric spectra approximately, however, it leads to a stronger dependence of the spin asymmetry on the modulation amplitude than cosine-type modulation does. These results show that phase modulation and ellipticity can control the momentum spectra patterns in vacuum pair production. Moreover, phase modulation can also affect the degree of spin asymmetry.

Introduction

Electron–positron pair production from vacuum is a typical nonperturbative process in strong-field quantum electrodynamics (QED) [1, 2, 3, 4, 5, 6]. The rate is exponentially suppressed below the Schwinger critical field Ecr≈1.3×1016 V/cmE_{\mathrm{cr}} \approx1.3 \times10^{16}\,\mathrm{V/cm} [7]. In time-dependent fields, multiphoton transitions allow pair production at subcritical field strengths [8]. The E-144 experiment at the Stanford Linear Accelerator Center provided an important demonstration of multiphoton strong-field QED [9]. High-intensity laser facilities have further stimulated studies of pair production in time-dependent fields [10, 11, 12, 13]. Several theoretical approaches have been developed, including quantum kinetic theory, the Dirac–Heisenberg–Wigner (DHW) formalism, worldline instantons, and other semiclassical methods [14, 15, 16, 17, 18, 19, 20, 21]. The temporal structure of the field can enhance pair production and produce interference effects in the momentum spectra [22, 23, 24, 25, 26, 27].

The momentum distributions of produced particles contain information about the temporal structure, polarization, and interference dynamics of the driving field [28, 29, 30, 31]. Interference between pair-production contributions at different times can produce oscillatory structures in the momentum spectra [27, 28]. Spiral and vortex structures have been studied in strong-field ionization and photodetachment [32, 33, 34, 35, 36]. Related structures appear in vacuum pair production driven by rotating and counter-rotating fields [37]. In bichromatic counter-rotating fields, the relative time delay controls the number of spiral arms [29, 30]. Frequency chirping also modifies interference effects and substantially reshapes the momentum spectra in circularly polarized fields [31]. These studies establish rotating electric fields as a suitable setting for investigating the interplay among field rotation, interference, and spin in pair production, and such fields have been analyzed using semiclassical and Wigner-function approaches [38, 39]. Circular polarization is particularly suitable for spin-resolved studies because the rotating field couples particle motion to spin dynamics [40]. In circularly polarized fields, spin-resolved momentum spectra with spiral structures have been reported [41], providing information on the production dynamics beyond spin-summed distributions [42].

Despite these developments, the combined effects of phase modulation and ellipticity on spin-resolved momentum spectra in spatially homogeneous elliptically polarized fields have not been systematically investigated. Phase modulation has been studied in spatially inhomogeneous fields [43], and pair production in fields with both spatial and temporal dependence has also been investigated [44, 45]. A linear chirp changes the carrier frequency monotonically. In contrast, sinusoidal phase modulation produces a periodically varying instantaneous frequency without changing the pulse envelope, thereby reshaping the interference pattern. Ellipticity determines the rotational character of the transverse field and adjusts the balance between transverse rotation and directional anisotropy. Because the two factors act simultaneously on the spectral geometry and on the spin precession, their combined influence cannot be inferred from either factor alone.

In this work, we study vacuum electron–positron pair production in phase-modulated elliptically polarized electric fields using the quantum kinetic equations (QKEs). We calculate spin-summed and spin-resolved momentum distributions for different modulation amplitudes, frequencies, ellipticities, and forms of phase modulation. For circular polarization, cosine-type modulation transforms ring-shaped distributions into spiral patterns. The deformation does not increase monotonically with the modulation amplitude or frequency. It peaks at intermediate values within the sampled range and is most clearly resolved in this case. Varying the ellipticity changes the angular deformation and the anisotropy of the momentum distribution. Sine-type modulation preserves ring-shaped, approximately mirror-symmetric distributions, but its spin asymmetry exhibits a much stronger dependence on the modulation amplitude compared with the cosine-type case. For circular polarization with δ=+1\delta= +1, the spin-down channel is favored throughout the ranges of the modulation amplitude and frequency considered. Reversing the field handedness reverses the favored channel while leaving the total yield unchanged.

This paper is organized as follows. Section II introduces the QKEs and the field model. Section III discusses how the momentum spectra, contour deformation, and spin-resolved yields and spin asymmetry depend on the modulation amplitude, frequency, and ellipticity. Section IV compares the momentum spectra and spin asymmetry for cosine-type and sine-type phase modulations. Section V summarizes the main results and provides an outlook.

Theoretical Framework

We consider vacuum electron–positron pair production in a spatially homogeneous time-dependent elliptically polarized electric field. Within the dipole approximation, the field is

E(t)=E0e−t2/(2τ2)1+δ2(cos⁡Φ(t)δsin⁡Φ(t)0)(1)\mathbf{E}(t)=\frac{E_0 e^{-t^2/(2\tau^2)}}{\sqrt{1+\delta^2}} \begin{pmatrix} \cos\Phi(t)\\ \delta\sin\Phi(t)\\ 0 \end{pmatrix} \tag*{(1)}

where E0E_0 is the field-strength parameter, τ\tau the pulse duration, and Φ(t)\Phi(t) the phase function. Here δ=0\delta=0 and ∣δ∣=1|\delta|=1 correspond to linear and circular polarization, respectively, and the sign of δ\delta sets the rotation sense. For δ>0\delta>0, the field rotates counterclockwise when viewed along the +z+z axis; δ<0\delta<0 corresponds to the opposite rotation sense. The factor 1/1+δ21/\sqrt{1+\delta^2} keeps the cycle-averaged intensity fixed as δ\delta varies. For 0≤∣δ∣≤10\leq|\delta|\leq1, the peak field amplitude is E0/1+δ2E_0/\sqrt{1+\delta^2}.

We consider two forms of phase modulation,

Φcos⁡(t)=bcos⁡(Ωt)+ω0t,Φsin⁡(t)=bsin⁡(Ωt)+ω0t,(2)\Phi_{\cos}(t)=b\cos(\Omega t)+\omega_0t,\qquad \Phi_{\sin}(t)=b\sin(\Omega t)+\omega_0t, \tag*{(2)}

where ω0\omega_0 is the carrier frequency, bb the modulation amplitude, and Ω\Omega the modulation frequency. The cosine-type modulation serves as the reference configuration, while the sine-type modulation is introduced to examine the role of the temporal phase structure. The corresponding instantaneous angular frequencies are

ωinstcos⁡(t)=dΦcos⁡dt=ω0−bΩsin⁡(Ωt),ωinstsin⁡(t)=dΦsin⁡dt=ω0+bΩcos⁡(Ωt).(3)\omega_{\mathrm{inst}}^{\cos}(t)=\frac{\mathrm{d}\Phi_{\cos}}{\mathrm{d}t}=\omega_0-b\Omega\sin(\Omega t),\qquad \omega_{\mathrm{inst}}^{\sin}(t)=\frac{\mathrm{d}\Phi_{\sin}}{\mathrm{d}t}=\omega_0+b\Omega\cos(\Omega t). \tag*{(3)}

To formulate the QKEs, we introduce the vector potential in the temporal gauge, A0=0A^0=0, for which

E(t)=−A˙(t).(4)\mathbf{E}(t)=-\dot{\mathbf{A}}(t). \tag*{(4)}

The integration constant in A(t)\mathbf{A}(t) is chosen according to the temporal symmetry of each field configuration. The kinetic momentum is

p(t)=q−eA(t).(5)\mathbf{p}(t)=\mathbf{q}-e\mathbf{A}(t). \tag*{(5)}

where q\mathbf{q} is the canonical momentum and ee is the elementary charge. We use natural units with ℏ=c=m=e=1\hbar=c=m=e=1.

The pair production process in a spatially homogeneous electric field can be described by the QKEs [17, 21], which are exactly equivalent to the DHW formalism in this limit [14, 15]. Following the standard reduction, the Wigner function is independent of the spatial coordinate, and only 10 equal-time components are nonvanishing. We organize these components as a scalar distribution ff, a spin vector F\mathbf{F}, and two coherence vectors u\mathbf{u} and v\mathbf{v}. The QKEs then read

f˙=−2μ2⋅u,(6)\dot{f}=-2\boldsymbol{\mu}_2\cdot\mathbf{u}, \tag*{(6)}
F˙=2μ1×F−2μ2×v,(7)\dot{\mathbf{F}}=2\boldsymbol{\mu}_1\times\mathbf{F}-2\boldsymbol{\mu}_2\times\mathbf{v}, \tag*{(7)}
u˙=2μ1×u+μ2(2f−1)+2εpv,(8)\dot{\mathbf{u}}=2\boldsymbol{\mu}_1\times\mathbf{u}+\boldsymbol{\mu}_2(2f-1)+2\varepsilon_p\mathbf{v}, \tag*{(8)}
v˙=2μ1×v−2μ2×F−2εpu,(9)\dot{\mathbf{v}}=2\boldsymbol{\mu}_1\times\mathbf{v}-2\boldsymbol{\mu}_2\times\mathbf{F}-2\varepsilon_p\mathbf{u}, \tag*{(9)}

where

μ1=p×E2εp(εp+1),μ2=p(E⋅p)−εp(εp+1)E2εp2(εp+1),(10)\boldsymbol{\mu}_1=\frac{\mathbf{p}\times\mathbf{E}}{2\varepsilon_p(\varepsilon_p+1)},\qquad \boldsymbol{\mu}_2=\frac{\mathbf{p}(\mathbf{E}\cdot\mathbf{p})-\varepsilon_p(\varepsilon_p+1)\mathbf{E}}{2\varepsilon_p^2(\varepsilon_p+1)}, \tag*{(10)}

with εp=p2+1\varepsilon_p=\sqrt{\mathbf{p}^2+1}. Here ff is the spin-averaged particle distribution and F\mathbf{F} carries the spin-dependent information. The initial conditions correspond to the vacuum, f=F=u=v=0f=\mathbf{F}=\mathbf{u}=\mathbf{v}=0.

The spin-resolved distributions are obtained from the covariant spin projection of the Wigner function [20, 39, 42]. We choose the zz axis perpendicular to the polarization plane as the spin-quantization axis, for which the two spin channels can have different number densities [42]. The spin-resolved momentum distributions at a sufficiently late time toutt_{\mathrm{out}} are

f↑(q)=f(q,tout)−Fz(q,tout),f↓(q)=f(q,tout)+Fz(q,tout),(11)f_{\uparrow}(\mathbf{q})=f(\mathbf{q},t_{\mathrm{out}})-F_z(\mathbf{q},t_{\mathrm{out}}),\qquad f_{\downarrow}(\mathbf{q})=f(\mathbf{q},t_{\mathrm{out}})+F_z(\mathbf{q},t_{\mathrm{out}}), \tag*{(11)}

where Fz>0F_z>0 corresponds to a larger spin-down population. Here f↑,↓f_{\uparrow,\downarrow} denote the spin-resolved distributions of the produced electrons. The positron spin projection is opposite along the same quantization axis. Their sum defines the spin-summed distribution,

ftotal(q)=f↑(q)+f↓(q)=2f(q,tout).(12)f_{\mathrm{total}}(\mathbf{q})=f_{\uparrow}(\mathbf{q})+f_{\downarrow}(\mathbf{q})=2f(\mathbf{q},t_{\mathrm{out}}). \tag*{(12)}

The momentum distributions below are evaluated at qz=0q_z=0 for the transverse spectra.

The resulting system of ordinary differential equations is solved numerically for the 10 Wigner components together with the three components of the vector potential. We use a variable-order Adams–Bashforth–Moulton solver with adaptive step size, setting the relative and absolute tolerances to 10−1410^{-14} and 10−1610^{-16}, respectively. The time integration runs from t=−10τt=-10\tau to t=10τt=10\tau, where the field has effectively vanished. The transverse momentum plane is sampled on a uniform 150×150150\times150 grid over qx,qy∈[−1.5m,1.5m]q_x,q_y\in[-1.5m,1.5m], corresponding to a grid spacing of Δq=0.02m\Delta q=0.02m. Test runs with a finer grid and a larger cutoff produced consistent results.

Spin-resolved momentum spectra and spin asymmetry

We first study vacuum electron–positron pair production for circular polarization, setting δ=1\delta= 1 and using the cosine-type phase modulation. The field-strength parameter, carrier frequency, and pulse duration are E0=0.12EcrE_0 = 0.1\sqrt{2}E_{\mathrm{cr}}, ω0=0.55 m\omega_0 = 0.55\,m, and τ=30/m\tau= 30/m, respectively. These parameters correspond to a subcritical field in the multiphoton regime and are of the same order as those used in previous studies [30, 37]. All frequencies and momenta below are given in units of mm.

Modulation amplitude bb

We fix Ω=0.04\Omega= 0.04 and vary b=1,2,6,9b = 1, 2, 6, 9. The spin-up momentum distributions are shown in Figure 1. For b=1b = 1, the distribution is nearly ring-shaped. As bb increases to 2 and 6, it develops a pronounced spiral pattern. At b=9b = 9, the distribution remains strongly deformed, but its contour becomes more complex.

Figure 1. Transverse momentum distributions of spin-up particles for different modulation amplitudes in a circularly polarized field with cosine-type phase modulation. The modulation amplitudes are b=1b = 1, 22, 66, and 99. The other parameters are E0=0.12EcrE_0 = 0.1\sqrt{2}E_{\mathrm{cr}}, ω0=0.55 m\omega_0 = 0.55\,m, τ=30/m\tau= 30/m, Ω=0.04\Omega= 0.04, and qz=0q_z = 0.

To characterize this deformation, we extract isodensity contours, which directly reflect the angular distortion of the momentum distribution. The contours are sampled over 150 isodensity levels equally spaced between 1% and 100% of the maximum density. The results from different levels are then combined through the weighted average. In polar coordinates, the radial distance rr and polar angle θ\theta are defined as

r=qx2+qy2,θ=arg⁡(qx+iqy),(13)r = \sqrt{q_x^2 + q_y^2}, \qquad\theta= \arg(q_x + iq_y), \tag*{(13)}

where qxq_x and qyq_y are the transverse momentum components. Using these coordinates, each extracted contour segment is fitted by

r(θ)=aiθ+ci,(14)r(\theta) = a_i\theta+ c_i, \tag*{(14)}

where aia_i approximates the local slope dr/dθdr/d\theta within the ii-th segment. The sign of aia_i gives the local direction of the radial change with the polar angle.

Segments with fewer than eight distinct polar-angle samples are discarded. For the remaining segments, a linear fit is retained only if its angular span Δθi\Delta\theta_i is nonzero and its coefficient of determination satisfies Ri2≥0.30R_i^2 \ge0.30. These thresholds exclude segments too short or too poorly resolved to define a reliable local slope, while retaining enough segments for a meaningful weighted average. After the eight-sample selection, the number of remaining segments is denoted by NsegN_{\mathrm{seg}}, and the number of segments retained after the linear-fit criteria is denoted by NfitN_{\mathrm{fit}}. The weighted mean slope is aˉ=∑iwiai/∑iwi\bar{a} = \sum_i w_i a_i / \sum_i w_i, with weights wi=ΔθiRi2w_i = \Delta\theta_i R_i^2, where the sum runs over retained segments. This weighting gives more weight to segments with larger angular spans and better linear fits. The weighted mean coefficient of determination is R2‾=∑iwiRi2/∑iwi\overline{R^2} = \sum_i w_i R_i^2 / \sum_i w_i. While aˉ\bar{a} characterizes the local radial change of the contour with respect to the polar angle, R2‾\overline{R^2} measures the average quality of the local linear fits. For convenience, we also list 2π∣aˉ∣2\pi|\bar{a}|, which estimates the total radial change of the contour over a full azimuthal turn. This procedure does not assume that the entire contour forms a single spiral. The local linear fitting results for the extracted contours are summa-

rized in Table I. The last column of Table I gives the sign of aˉ\bar{a}: −-, ++, and 0 correspond to dr/dθ<0dr/d\theta< 0, dr/dθ>0dr/d\theta> 0, and a negligibly small ∣aˉ∣|\bar{a}|, respectively.

bbNsegN_{\mathrm{seg}}NfitN_{\mathrm{fit}}aˉ\bar{a}R2‾\overline{R^2}2π∣aˉ∣2\pi|\bar{a}|Sign
15311364.73×10−54.73 \times10^{-5}0.45592.97×10−42.97 \times10^{-4}0
2415226−1.20×10−2-1.20 \times10^{-2}0.60467.52×10−27.52 \times10^{-2}−-
6252213−3.37×10−2-3.37 \times10^{-2}0.79282.12×10−12.12 \times10^{-1}−-
9233202−3.28×10−2-3.28 \times10^{-2}0.64242.06×10−12.06 \times10^{-1}−-

Table I. Local linear contour-fitting results for different modulation amplitudes.

The weighted mean slope is close to zero for b=1b = 1 and becomes clearly negative at b=2b = 2. Both ∣aˉ∣|\bar{a}| and 2π∣aˉ∣2\pi|\bar{a}| are largest at b=6b = 6, where R2‾\overline{R^2} is also highest. At b=9b = 9, the deformation remains strong, but the lower R2‾\overline{R^2} is consistent with a more complex contour geometry. The deformation is therefore most pronounced at b=6b = 6 in the range considered and does not increase monotonically with bb. At fixed Ω\Omega, increasing bb increases ∣bΩ∣|b\Omega|, the amplitude of the instantaneous-frequency variation, so particles produced at different times acquire different final momenta. At larger bb, interference effects between contributions from different production times generate contour segments with distinct local orientations, which results in more complex structures and poorer linear fits.

The spin-down distributions for the same field parameters are shown in Figure 2. Both spin channels exhibit qualitatively similar deformations, indicating that the spiral structure is a property of the pair-production dynamics rather than of a specific spin channel. The spiral pattern is somewhat more pronounced in the spin-up channel, so we use the spin-up distributions for the contour analysis in the following subsections.

Transverse momentum distribution for $b=1$

(a) b=1b = 1

Transverse momentum distribution for $b=2$

(b) b=2b = 2

Transverse momentum distribution for $b=6$

(c) b=6b = 6

Transverse momentum distribution for $b=9$

(d) b=9b = 9

Figure 2. Transverse momentum distributions of spin-down particles for different modulation amplitudes in a circularly polarized field with cosine-type phase modulation. The parameters are the same as in Figure 1.

Modulation frequency Ω\Omega

We fix b=5b = 5 and vary Ω=0.02,0.05,0.06,0.08\Omega= 0.02, 0.05, 0.06, 0.08. The other parameters are the same as in Figure 1. The spin-up momentum distributions are shown in Figure 3. At Ω=0.02\Omega= 0.02, the distribution is close to a ring with only weak deformation. As Ω\Omega increases from 0.02 to 0.05, the ring becomes more distorted and a spiral structure develops. At Ω=0.06\Omega= 0.06, the deformation weakens. At Ω=0.08\Omega= 0.08, the net orientation of the contour slopes reverses.

Figure 3. Transverse momentum distributions of spin-up particles for different modulation frequencies in a circularly polarized field with cosine-type phase modulation. The modulation frequencies are Ω=0.02,0.05,0.06,0.08\Omega= 0.02, 0.05, 0.06, 0.08, with b=5b = 5 and qz=0q_z = 0. The other parameters are the same as in Figure 1.

Ω\OmegaNsegN_{\mathrm{seg}}NfitN_{\mathrm{fit}}aˉ\bar{a}R2‾\overline{R^2}2π∣aˉ∣2\pi|\bar{a}|Sign
0.02476173−1.71×10−3-1.71 \times10^{-3}0.50021.08×10−21.08 \times10^{-2}0
0.05213178−2.51×10−2-2.51 \times10^{-2}0.53381.58×10−11.58 \times10^{-1}−-
0.06221124−1.74×10−2-1.74 \times10^{-2}0.42671.09×10−11.09 \times10^{-1}−-
0.083071918.72×10−38.72 \times10^{-3}0.53505.48×10−25.48 \times10^{-2}++

Table 2. Local linear contour-fitting results for different modulation frequencies.

We use the same local contour-fitting procedure as in Sec. III A for the spin-up distributions. The fitting results are listed in Table II. The contour deformation varies nonmonotonically with Ω\Omega. At Ω=0.02\Omega= 0.02, the weighted mean slope is small, consistent with the nearly ring-shaped distribution. The deformation is strongest at Ω=0.05\Omega= 0.05 and weaker at Ω=0.06\Omega= 0.06, where the lower Rˉ2\bar{R}^{2} indicates that the contours are less well described by local linear fits. At Ω=0.08\Omega= 0.08, the sign of aˉ\bar{a} changes, indicating a change in the net direction inferred from the local fits. Varying Ω\Omega at fixed bb changes the amplitude ∣bΩ∣|b\Omega| of the instantaneous-frequency variation. Increasing ∣bΩ∣|b\Omega| makes the instantaneous frequency vary more strongly during the pulse, so particles produced at different times accumulate different phases. This changes the angular distribution of the final momenta and initially enhances the deformation. At sufficiently large ∣bΩ∣|b\Omega|, however, the rapid phase variation can modify the interference pattern and even reverse the net direction of the local contour slopes.

When bb is varied with Ω=0.04\Omega= 0.04, the most pronounced spiral appears at b=6b = 6, giving ∣bΩ∣=0.24|b\Omega| = 0.24. When Ω\Omega is varied with b=5b = 5, the strongest deformation occurs at Ω=0.05\Omega= 0.05, giving ∣bΩ∣=0.25|b\Omega| = 0.25. The similar values of ∣bΩ∣|b\Omega| at the peak deformation suggest that this product is a key parameter determining the maximum deformation. Beyond this common dependence, Ω\Omega also changes the modulation period 2π/Ω2\pi/\Omega. Since the pulse duration is fixed, different Ω\Omega values correspond to different fractions of a modulation cycle being sampled within the pulse. This difference in the sampled phase range can reverse the net direction of the local contour slopes at Ω=0.08\Omega= 0.08, whereas no such reversal appears over the range of bb considered here. Thus, Ω\Omega provides an additional control parameter for the orientation of the spiral deformation.

Ellipticity $\delta`

We fix b=6b = 6 and Ω=0.04\Omega= 0.04, for which the spiral deformation was most pronounced in Sec. III A, and vary δ=0.2,0.4,0.6,0.8,1.0\delta= 0.2, 0.4, 0.6, 0.8, 1.0. The remaining parameters are the same as in Figure 1. The spin-up momentum distributions are shown in Figure 4.

Transverse momentum distributions of spin-up particles for different ellipticities in a cosine-type phase-modulated field

Figure 4. Transverse momentum distributions of spin-up particles for different ellipticities in a cosine-type phase-modulated field. The ellipticities are δ=0.2\delta= 0.2, 0.40.4, 0.60.6, 0.80.8, and 1.01.0. The modulation parameters are b=6b = 6 and Ω=0.04\Omega= 0.04, with qz=0q_z = 0. The other parameters are E0=0.12EcrE_0 = 0.1\sqrt{2}E_{\mathrm{cr}}, ω0=0.55m\omega_0 = 0.55m, and τ=30/m\tau= 30/m.

For small δ\delta, the distribution is strongly anisotropic and elongated along the qxq_x direction, consistent with the nearly linear polarization of the field [46]. As δ\delta increases, the overall distribution becomes less elongated and more ring-like, while the contour develops a spiral deformation.

To quantify the contour deformation, we use the local contour-fitting procedure described above. The fitting results are listed in Table III. For elliptical fields, aˉ\bar{a} is not a robust global descriptor because the contours are locally anisotropic, and NfitN_{\mathrm{fit}} is much smaller. The circularly polarized case has the largest NfitN_{\mathrm{fit}} and the best mean fit quality, so it gives the most reliable contour description. In particular, only six segments are retained at δ=0.8\delta= 0.8, so the corresponding aˉ\bar{a} and R2‾\overline{R^2} values are statistically limited and should not be overinterpreted.

δ\deltaNsegN_{\mathrm{seg}}NfitN_{\mathrm{fit}}aˉ\bar{a}R2‾\overline{R^2}2π∣aˉ∣2\pi|\bar{a}|Sign
0.2205271.356×10−21.356 \times10^{-2}0.43818.52×10−28.52 \times10^{-2}+
0.423034−6.620×10−3-6.620 \times10^{-3}0.46364.16×10−24.16 \times10^{-2}−
0.632918−2.626×10−2-2.626 \times10^{-2}0.40871.65×10−11.65 \times10^{-1}−
0.845661.690×10−21.690 \times10^{-2}0.59581.06×10−11.06 \times10^{-1}+
1.0252213−3.369×10−2-3.369 \times10^{-2}0.79282.12×10−12.12 \times10^{-1}−

Table III. Local linear contour-fitting results for different ellipticities.

As noted above, the normalization keeps the cycle-averaged intensity fixed while reducing the peak amplitude. The amplitude ratio Ey/ExE_y/E_x is determined by δ\delta. As δ\delta increases, the yy-component of the transverse motion becomes stronger, and the overall motion becomes more rotational. At δ=1.0\delta= 1.0, the two transverse components have equal amplitudes, so the field has no preferred transverse axis. This makes the transverse motion more isotropic, so the modulation-induced phase differences can directly modify the azimuthal distribution, making the spiral pattern most pronounced among the sampled ellipticities. These results indicate that the spiral structure is governed jointly by the phase modulation and the ellipticity. The phase modulation controls the temporal phase evolution and the instantaneous frequency, while the ellipticity controls the balance between transverse rotation and directional anisotropy. Ellipticity thus provides an additional control parameter for the geometry of the transverse momentum spectra.

Spin-resolved yields and spin asymmetry

We analyze the spin-resolved yields and spin asymmetry in the qz=0q_z = 0 transverse momentum plane. The spin-up and spin-down yields are N↑,↓=∫dqx dqy f↑,↓(qx,qy,qz=0)N_{\uparrow,\downarrow} = \int dq_x\,dq_y\,f_{\uparrow,\downarrow}(q_x,q_y,q_z = 0). The total yield in this plane is Ntotal=N↑+N↓N_{\mathrm{total}} = N_{\uparrow} + N_{\downarrow}. We characterize the spin-resolved yields by the spin asymmetry and the spin-up fraction,

Ktotal=N↑−N↓N↑+N↓,P↑=N↑N↑+N↓×100%.(15)\mathcal{K}_{\mathrm{total}} = \frac{N_{\uparrow} - N_{\downarrow}}{N_{\uparrow} + N_{\downarrow}}, \qquad P_{\uparrow} = \frac{N_{\uparrow}}{N_{\uparrow} + N_{\downarrow}} \times100\%. \tag*{(15)}

With this convention, Ktotal>0\mathcal{K}_{\mathrm{total}} > 0 corresponds to N↑>N↓N_{\uparrow} > N_{\downarrow}, and Ktotal<0\mathcal{K}_{\mathrm{total}} < 0 corresponds to N↓>N↑N_{\downarrow} > N_{\uparrow}.

The dependence on bb is shown in Figure 5. The spin-resolved yields are shown in Figure 5a, and the corresponding spin asymmetry and spin-up fraction are shown in Figure 5b. The spin-up and total yields increase with bb over the considered range, while the spin-down yield remains substantially larger than the spin-up yield. For δ=+1\delta= +1, the spin asymmetry varies only weakly and nonmonotonically with bb, while the spin-down channel remains favored for all bb.

Spin-resolved yields and spin asymmetry

Figure 5. Spin-resolved yields and spin asymmetry as functions of the modulation amplitude bb in a circularly polarized field with cosine-type phase modulation. The modulation frequency is fixed at Ω=0.04\Omega= 0.04; the other parameters are the same as in Figure 1.

The dependence on Ω\Omega is shown in Figure 6. The spin-resolved yields are shown in Figure 6a, and the corresponding spin asymmetry and spin-up fraction are shown in Figure 6b. Both the spin-up and total yields grow with Ω\Omega over the considered range. The spin-down yield provides the majority of the total yield, while the spin-up yield is considerably smaller. For δ=+1\delta= +1, the degree of the spin asymmetry varies weakly and nonmonotonically with Ω\Omega, while the spin-down channel remains favored for all Ω\Omega.

Spin-resolved yields and spin asymmetry

Figure 6. Spin-resolved yields and spin asymmetry as functions of the modulation frequency Ω\Omega in a circularly polarized field with cosine-type phase modulation. The modulation amplitude is fixed at b=5b=5; the other parameters are the same as in Figure 1.

For δ=+1\delta= +1, the modulation amplitude and frequency affect the degree of the spin asymmetry but not the favored spin channel, which remains spin-down. The asymmetry originates from the field rotation, the time-dependent phase modulation, and the coupled spin dynamics discussed below. The total yield and the spin asymmetry show different trends as bb and Ω\Omega vary, suggesting that pair production and spin imbalance are not simply proportional.

To examine whether the favored spin channel is tied to the field handedness, we repeated the calculation with δ=−1\delta= -1. Table IV shows that at Ω=0.04\Omega= 0.04 the favored channel reverses for both cosine-type and sine-type modulations, with N↑(δ=−1)=N↓(δ=+1)N_{\uparrow}(\delta= -1) = N_{\downarrow}(\delta= +1) and N↓(δ=−1)=N↑(δ=+1)N_{\downarrow}(\delta= -1) = N_{\uparrow}(\delta= +1) to numerical precision, while the total yield remains unchanged at fixed bb and Ω\Omega. The same reversal is found for the cosine-type modulation at Ω=0.02\Omega= 0.02, 0.050.05, 0.060.06, and 0.080.08. The magnitude ∣Ktotal∣|K_{\mathrm{total}}| stays between 0.960.96 and 0.990.99 in all cases.

ModulationbbKtotal(δ=+1)K_{\mathrm{total}}(\delta= +1)Ktotal(δ=−1)K_{\mathrm{total}}(\delta= -1)
Cosine1−9.883×10−1-9.883 \times10^{-1}+9.883×10−1+9.883 \times10^{-1}
Cosine2−9.858×10−1-9.858 \times10^{-1}+9.858×10−1+9.858 \times10^{-1}
Cosine6−9.888×10−1-9.888 \times10^{-1}+9.888×10−1+9.888 \times10^{-1}
Cosine9−9.838×10−1-9.838 \times10^{-1}+9.838×10−1+9.838 \times10^{-1}
Sine1−9.856×10−1-9.856 \times10^{-1}+9.856×10−1+9.856 \times10^{-1}
Sine2−9.759×10−1-9.759 \times10^{-1}+9.759×10−1+9.759 \times10^{-1}
Sine6−9.879×10−1-9.879 \times10^{-1}+9.879×10−1+9.879 \times10^{-1}
Sine9−9.652×10−1-9.652 \times10^{-1}+9.652×10−1+9.652 \times10^{-1}

Table 4. Spin asymmetry KtotalK_{\mathrm{total}} for opposite field handedness at Ω=0.04\Omega= 0.04.

We further qualitatively interpret the spin asymmetry using the effective magnetic field of a moving particle. In the QKEs, the spin-vector evolution is F˙=2μ1×F−2μ2×v\dot{\mathbf{F}}=2\boldsymbol{\mu}_1 \times\mathbf{F}-2\boldsymbol{\mu}_2 \times\mathbf{v}, with μ1=p×E/[2εp(εp+1)]\boldsymbol{\mu}_1=\mathbf{p}\times\mathbf{E}/[2\varepsilon_p(\varepsilon_p+1)]. Defining Beff≡2μ1\mathbf{B}_{\mathrm{eff}}\equiv2\boldsymbol{\mu}_1, the spin-precession equation becomes F˙=Beff×F−2μ2×v\dot{\mathbf{F}}=\mathbf{B}_{\mathrm{eff}}\times\mathbf{F}-2\boldsymbol{\mu}_2\times\mathbf{v}. The source term −2μ2×v-2\boldsymbol{\mu}_2\times\mathbf{v} excites FzF_z from zero, while Beff\mathbf{B}_{\mathrm{eff}} drives the spin precession. Their interplay determines the momentum-space structure of Fz(q)F_z(\mathbf{q}) and hence the spin asymmetry.

In the nonrelativistic limit, the effective field reduces to Beff→12vcl×E\mathbf{B}_{\mathrm{eff}}\to\frac{1}{2}\mathbf{v}_{\mathrm{cl}}\times\mathbf{E}, where vcl=p/εp\mathbf{v}_{\mathrm{cl}}=\mathbf{p}/\varepsilon_p is the classical velocity. This differs from the motional field −vcl×E-\mathbf{v}_{\mathrm{cl}}\times\mathbf{E} by a factor of −1/2-1/2, which reflects the relativistic Thomas contribution in the precession rate of the QKEs. This result is consistent with the difference between the Foldy–Wouthuysen expansion and the leading-order Zeeman shift [47]. The effective field couples to the spin through HZ≃−μ⋅BeffH_Z\simeq-\boldsymbol{\mu}\cdot\mathbf{B}_{\mathrm{eff}}, where μ\boldsymbol{\mu} is the spin magnetic moment. Because the spin magnetic moment μ\boldsymbol{\mu} has opposite projections along Beff\mathbf{B}_{\mathrm{eff}} for spin-up and spin-down states, the two states experience opposite energy shifts. Since Beff\mathbf{B}_{\mathrm{eff}} depends on momentum, particles with different momenta accumulate different spin phases and thus different production probabilities. The favored spin channel is determined by the field handedness, which remains unchanged under variations of bb and Ω\Omega in the parameter range considered. This is consistent with the δ=±1\delta=\pm1 comparison in Table IV.

Effects of the phase-modulation form

We now study how the form of the phase modulation affects the momentum distributions and spin asymmetry, comparing the cosine-type and sine-type modulations. The remaining parameters are the same as in Sec. III. The sine-type results for the spin-summed, spin-up, and spin-down distributions are shown in Figure 7. The corresponding cosine-type spin-up distributions were discussed in Sec. III A.

Transverse momentum distributions of spin-summed, spin-up, and spin-down particles for the sine-type phase modulation

Figure 7. Transverse momentum distributions of spin-summed, spin-up, and spin-down particles for the sine-type phase modulation in a circularly polarized field. The modulation frequency is Ω=0.04\Omega= 0.04, the modulation amplitudes are b=1b = 1, 2, 6, and 9, and qz=0q_z = 0. The remaining parameters are the same as in Figure 1.

For the sine-type modulation, the three distributions remain predominantly ring-shaped and approximately mirror-symmetric about the qyq_y axis as bb increases. The two spin-resolved distributions have similar momentum-space patterns but different magnitudes. In contrast, the cosine-type spin-up distributions show stronger directional deformation and a more pronounced spiral structure, particularly at larger bb. This difference originates from the temporal symmetry of the field. For the sine-type modulation, Ex(t)E_x(t) is even and Ey(t)E_y(t) is odd under t→−tt \to-t. With the integration constant chosen such that Ax(0)=Ay(0)=0A_x(0) = A_y(0) = 0, the vector potential satisfies Ax(−t)=−Ax(t)A_x(-t) = -A_x(t) and Ay(−t)=Ay(t)A_y(-t) = A_y(t). Under the transformation t→−tt \to-t together with qx→−qxq_x \to-q_x, the term qx−eAx(t)q_x - eA_x(t) changes sign while qy−eAy(t)q_y - eA_y(t) is unchanged, so the instantaneous energy ε(q,t)=1+[qx−eAx(t)]2+[qy−eAy(t)]2\varepsilon(\mathbf{q},t) = \sqrt{1 + [q_x - eA_x(t)]^2 + [q_y - eA_y(t)]^2} remains invariant. This is consistent with the approximately mirror-symmetric distribution about the qyq_y axis. For the cosine-type modulation, these field components are neither even nor odd, so this symmetry is broken and the distribution can develop stronger directional deformation and spiral structure.

We now compare the spin asymmetry produced by the two modulation forms. The dependence of Ktotal\mathcal{K}_{\mathrm{total}} on bb is shown in Figure 8(a). As shown in Sec. III D, the spin asymmetry for the cosine-type modulation varies only weakly and nonmonotonically with bb. In contrast, the sine-type modulation shows a stronger dependence on bb: the spin-down dominance is weaker at b=1b = 1 and b=2b = 2, remains strong at b=6b = 6, and decreases again at b=9b = 9. The cosine-type modulation strongly affects the directional structure while leaving the asymmetry stable, whereas the sine-type modulation preserves the mirror-symmetric pattern but makes the asymmetry strongly dependent on bb. The phase-modulation form therefore affects the spin-resolved observables through both the momentum-space geometry and the relative weighting of the two spin channels. Reversing the field handedness reverses the favored channel for the sine-type modulation as well, with the total yield unchanged (Table IV). The handedness–spin correspondence is therefore a property of the circularly polarized field itself and does

Figure 8. Spin asymmetry and its difference for cosine-type and sine-type phase modulations in a circularly polarized field with Ω=0.04\Omega= 0.04. (a) The total spin asymmetry Ktotal\mathcal{K}_{\mathrm{total}} as a function of the modulation amplitude bb. (b) The difference ΔKtotal=Ktotalsin−Ktotalcos\Delta\mathcal{K}_{\mathrm{total}} = \mathcal{K}^{\mathrm{sin}}_{\mathrm{total}} - \mathcal{K}^{\mathrm{cos}}_{\mathrm{total}} versus bb, where red and blue bars correspond to positive and negative values, respectively, and the shading intensity reflects their magnitude.

not depend on the modulation form. The difference between the two forms is quantified by

ΔKtotal=Ktotalsin−Ktotalcos,(16)\Delta\mathcal{K}_{\mathrm{total}} = \mathcal{K}^{\mathrm{sin}}_{\mathrm{total}} - \mathcal{K}^{\mathrm{cos}}_{\mathrm{total}}, \tag*{(16)}

shown in Figure 8(b). The difference changes sign as bb varies: it is positive for b=1,2,3b = 1, 2, 3, negative near b=4b = 4 and b=5b = 5, and positive again at b=6b = 6 and b=9b = 9. Neither form consistently produces a larger ∣Ktotal∣\lvert\mathcal{K}_{\mathrm{total}}\rvert over the entire parameter range, and their relative magnitudes depend sensitively on bb.

The nonmonotonic behavior of ΔKtotal\Delta\mathcal{K}_{\mathrm{total}} reflects the different temporal phase structures of the two modulations. Both fields have the same carrier frequency, pulse envelope, field strength, and ellipticity. However, the sine- and cosine-type modulations place the strong-field region at different phases of the field rotation. This modifies the coupling among the field, the particle motion, and the spin dynamics. The spin asymmetry is therefore sensitive not only to the overall field strength but also to the temporal phase structure of the pulse. The difference originates from the different time evolutions of the effective magnetic field Beff=2μ1\mathbf{B}_{\mathrm{eff}} = 2\mu_{1} experienced by a moving particle. The sine- and cosine-type modulations produce different instantaneous frequencies ωinst(t)\omega_{\mathrm{inst}}(t), and hence different field histories E(t)\mathbf{E}(t) and different Beff(t)\mathbf{B}_{\mathrm{eff}}(t) throughout the pulse.

Summary and Outlook

We have studied vacuum electron–positron pair production in phase-modulated elliptically polarized electric fields using the QKEs, examining how the modulation amplitude, frequency, ellipticity, and form of phase modulation affect the spin-resolved momentum spectra and the spin asymmetry. For circular polarization with cosine-type modulation, increasing the modulation amplitude deforms ring-shaped distributions into spiral patterns. The deformation does not increase monotonically with either the modulation amplitude or the frequency, and is most pronounced at intermediate values within the sampled range. At larger amplitudes, the contour develops segments with different local orientations, and the local linear fits become less reliable. As the ellipticity increases, the distribution becomes less elongated and more ring-like, and the spiral deformation becomes more pronounced; at δ=1\delta= 1, the transverse motion is most isotropic, so the modulation-induced phase differences directly reshape the azimuthal distribution.

The form of the phase modulation affects the spin asymmetry independently of the momentum-space geometry. For circular polarization with δ=+1\delta= +1, the spin-down channel is favored over the ranges of modulation amplitude and frequency considered in the qz=0q_z = 0 plane. Reversing the field handedness reverses the favored channel and leaves the total yield unchanged, for both cosine-type and sine-type modulations. Sine-type modulation keeps the distribution ring-shaped and approximately mirror-symmetric, but makes the asymmetry strongly dependent on the modulation amplitude. Since both modulations share the same carrier frequency, pulse envelope, field strength, and ellipticity, the difference must originate from their temporal phase structure, which places the strong-field region at different phases of the field rotation.

Ellipticity thus sets the overall shape of the momentum distribution, while phase modulation controls the spiral structure and the degree of spin asymmetry. Because geometric deformation and spin imbalance are governed by different aspects of the field, they can be analyzed separately in spin-resolved studies. The present analysis does not explain the spiral structures dynamically. A complex-time semiclassical analysis could identify the dominant turning points and clarify the interference mechanisms behind the observed momentum patterns [48].

Two natural extensions are to consider the involved problem with phase modulation in spatially inhomogeneous fields and magnetic fields as well. Spatially inhomogeneous fields introduce additional geometric effects through the finite extent of realistic laser pulses, while magnetic fields couple directly to the spin and may change both the momentum spectra and the spin asymmetry. These problems, which are beyond present work, would be interesting and worthy to be studied in future furthermore.

Acknowledgments

This work was supported by the National Natural Science Foundation of China under Grant Nos. 12375240 and 11935008. The computation was carried out at the HSCC of Beijing Normal University.

References

  1. [1]F. Sauter, Z. Phys. 69, 742 (1931).
  2. [2]W. Heisenberg and H. Euler, Z. Phys. 98, 714 (1936).
  3. [3]J. Schwinger, Phys. Rev. 82, 664 (1951).
  4. [4]A. Di Piazza, C. Müller, K. Z. Hatsagortsyan, and C. H. Keitel, Rev. Mod. Phys. 84, 1177 (2012).
  5. [5]A. Fedotov, A. Ilderton, F. Karbstein, B. King, D. Seipt, H. Taya, and G. Torgrimsson, Phys. Rep. 1010, 1 (2023).
  6. [6]B. S. Xie, Z. L. Li, and S. Tang, Matter Radiat. Extremes 2, 225 (2017).
  7. [7]M. S. Marinov and V. S. Popov, Fortschr. Phys. 25, 373 (1977).
  8. [8]E. Brézin and C. Itzykson, Phys. Rev. D 2, 1191 (1970).
  9. [9]D. L. Burke, R. C. Field, G. Horton-Smith, J. E. Spencer, D. Walz, S. C. Berridge, W. M. Bugg, K. Shmakov, A. W. Weidemann, C. Bula, et al., Phys. Rev. Lett. 79, 1626 (1997).
  10. [10]G. A. Mourou, T. Tajima, and S. V. Bulanov, Rev. Mod. Phys. 78, 309 (2006).
  11. [11]I. C. E. Turcu, S. Balascuta, F. Negoita, D. Jaroszynski, and P. McKenna, AIP Conf. Proc. 1645, 416 (2015).
  12. [12]I. Y. Kostyukov, E. A. Khazanov, A. A. Shaikin, A. G. Litvak, and A. M. Sergeev, Bull. Lebedev Phys. Inst. 50, S635 (2023).
  13. [13]A. Ringwald, Phys. Lett. B 510, 107 (2001).
  14. [14]D. Vasak, M. Gyulassy, and H. T. Elze, Ann. Phys. 173, 462 (1987).
  15. [15]I. Bialynicki-Birula, P. Gornicki, and J. Rafelski, Phys. Rev. D 44, 1825 (1991).
  16. [16]I. Bickers and N. S. Witte, Phys. Rev. D 89, 036002 (2014).
  17. [17]F. Hebenstreit, R. Alkofer, and H. Gies, Phys. Rev. D 82, 105026 (2010).
  18. [18]G. V. Dunne, Q. H. Wang, H. Gies, and C. Schubert, Phys. Rev. D 73, 065028 (2006).
  19. [19]G. V. Dunne and Q. H. Wang, Phys. Rev. D 74, 065015 (2006).
  20. [20]A. Blinne and E. Strobel, Phys. Rev. D 93, 025014 (2016).
  21. [21]I. A. Aleksandrov, A. Kudlis, and A. I. Klochai, Phys. Rev. Res. 6, 043009 (2024).
  22. [22]R. Schützhold, H. Gies, and G. Dunne, Phys. Rev. Lett. 101, 130404 (2008).
  23. [23]G. V. Dunne, H. Gies, and R. Schützhold, Phys. Rev. D 80, 111301 (2009).
  24. [24]M. F. Linder, C. Schneider, J. Sicking, N. Szpak, and R. Schützhold, Phys. Rev. D 92, 085009 (2015).
  25. [25]A. Otto, D. Seipt, D. Blaschke, B. Kämpfer, and S. A. Smolyansky, Phys. Lett. B 740, 335 (2015).
  26. [26]C. K. Dumlu and G. V. Dunne, Phys. Rev. Lett. 104, 250402 (2010).
  27. [27]E. Akkermans and G. V. Dunne, Phys. Rev. Lett. 108, 030401 (2012).
  28. [28]C. K. Dumlu and G. V. Dunne, Phys. Rev. D 83, 065028 (2011).
  29. [29]Z. L. Li, Y. J. Li, and B. S. Xie, Phys. Rev. D 96, 076010 (2017).
  30. [30]L. N. Hu, O. Amat, L. Wang, A. Sawut, H. H. Fan, and B. S. Xie, Phys. Rev. D 107, 116010 (2023).
  31. [31]O. Olugh, Z. L. Li, B. S. Xie, and R. Alkofer, Phys. Rev. D 99, 036003 (2019).
  32. [32]J. H. Macek, J. B. Sternberg, S. Y. Ovchinnikov, T. G. Lee, and D. R. Schultz, Phys. Rev. Lett. 102, 143201 (2009).
  33. [33]J. M. Ngoko Djiokap, S. X. Hu, L. B. Madsen, N. L. Manakov, A. V. Meremianin, and A. F. Starace, Phys. Rev. Lett. 115, 113004 (2015).
  34. [34]D. Pengel, S. Kerbstadt, D. Johannmeyer, L. Englert, T. Bayer, and M. Wollenhaupt, Phys. Rev. Lett. 118, 053003 (2017).
  35. [35]L. Ji, W. W. Yu, S. Y. Han, Y. F. Han, Q. Zhen, S. Q. Zhang, and X. S. Liu, Chem. Phys. Lett. 754, 137634 (2020).
  36. [36]M. M. Majczak, F. Cajiao Vélez, J. Z. Kamiński, and K. Krajewska, Opt. Express 30, 43330 (2022).
  37. [37]L. N. Hu, H. H. Fan, O. Amat, S. Tang, and B. S. Xie, Phys. Rev. D 110, 056013 (2024).
  38. [38]A. Blinne and H. Gies, Phys. Rev. D 89, 085001 (2014).
  39. [39]E. Strobel and S. S. Xue, Phys. Rev. D 91, 045016 (2015).
  40. [40]C. Kohlfürst, Phys. Rev. D 99, 096017 (2019).
  41. [41]Z. Y. Chen, O. Amat, J. H. Bai, and M. A. Bake, Phys. Rev. D 111, 116005 (2025).
  42. [42]R. Z. Jiang, Z. L. Li, and Y. J. Li, Phys. Rev. D 113, 116009 (2026).
  43. [43]L. N. Hu, O. Amat, L. J. Li, M. Mohamedsedik, and B. S. Xie, Commun. Theor. Phys. 75, 025102 (2023).
  44. [44]M. Ruf, G. R. Mocken, C. Müller, K. Z. Hatsagortsyan, and C. H. Keitel, Phys. Rev. Lett. 102, 080402 (2009).
  45. [45]F. Hebenstreit, R. Alkofer, and H. Gies, Phys. Rev. Lett. 107, 180403 (2011).
  46. [46]O. Olugh, Z. L. Li, and B. S. Xie, High Power Laser Sci. Eng. 8, e38 (2020).
  47. [47]A. Wöllert, H. Bauke, and C. H. Keitel, Phys. Rev. D 91, 125026 (2015).
  48. [48]C. K. Dumlu and G. V. Dunne, Phys. Rev. D 84, 125023 (2011).

Paper details

vlnc:W_kz2x2ctt
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