betaFeedback

WorksPhysics and Astronomy

Electron beam polarimeter and energy spectrometer

Abstract

The backscattering of laser radiation by a relativistic electron beam is a widely used method for measuring the spin polarization of electrons. We consider again the properties of the scattering process paying special attention to recoil electrons. Based on this consideration we propose the concept of the Compton polarimeter in which, in addition to all the polarization components, it becomes possible to accurately measure the energy and other parameters of the electron beam. To demonstrate the capabilities of the method we conduct a Monte Carlo simulations of the polarimeter developed for the FCC-ee project.

Inverse Compton scattering

Shortly after the invention of lasers it was pointed out in refs. [1, 2] that Inverse Compton Scattering (ICS) of laser light by relativistic electron beam would produce useful yields of high energy polarized photons. In 1969 such a beam had been realized at SLAC for a study of γp\gamma p interactions [3]. The effect of self-polarization of relativistic electrons due to emission of the spin-flip synchrotron radiation was predicted in ref. [4] and observed at the first ever e+e−\mathrm{e}^{+}\mathrm{e}^{-} collider ACO [5] due to polarization effects in intra-beam scattering. Soon the existence of this phenomenon was confirmed on the storage rings VEPP-2 and SPEAR as described in review [6]. Theoretical and experimental studies of spin dynamics in storage rings made it possible to develop a method of resonant depolarization described e. g. in ref. [7] and references therein. For the first time it was used for the beam energy calibration in experiments on precise determination of the Φ\Phi and KK mesons masses [8]. At higher beam energies in addition to other drawbacks the intra-beam scattering ceases to be a good tool for polarization measurement. ICS was proposed in ref. [9] as one of the processes allowing to measure the transverse spin polarization of high-energy electron beam. At the spin polarimeters of the VEPP-4 [10] and VEPP-2M [11] colliders the synchrotron or undulator radiation was scattered in the e+e−\mathrm{e}^{+}\mathrm{e}^{-} collision area. ICS of laser light for transverse beam polarization measurement was first used at SPEAR [12] and then at circular colliders PETRA [13], DORIS-II [14], LEP [15, 16], HERA [17] and VEPP-4M [18]. The ICS for longitudinal spin polarimetry was realized at SLAC linear collider [19], AmPS ring at NIKHEF [20], HERA collider at DESY [21] and Jlab [22]. The ICS polarimeters are also planned for future collider experiments ILC [23, 24], FCC-ee [25], CEPC [26, 27] and EIC [28, 29].

The practical realization of the ICS looks like it is shown in Figure 1. Laser radiation is inserted into the accelerator vacuum chamber and focused to the interaction point. The plane of the figure is the plane of machine, the zz-axis of the curvilinear coordinate system is always directed along the electron beam momentum. The xx-axis is shown in Figure 1 while the yy-axis is perpendicular to the plane of machine. The dipole magnet separates the scattered photons from the electron beam. The beam is bent by the angle θ0\theta_0 and the recoil electrons are deflected by a larger angle θ0+Δθ\theta_0 + \Delta\theta.

Image not extracted

Figure 1. Layout of ICS experiment.

Angle θ0\theta_0 depends on the ratio of the bending field integral to the electron energy ε0=γmc2\varepsilon_0 = \gamma mc^2:

θ0=∫B⊥ dlγ(mc/e),wheremce=1.704509024⋅10−3 [T m].(1)\theta_0 = \frac{\int B_{\perp}\,\mathrm{d}l}{\gamma(mc/e)}, \qquad\text{where} \quad\frac{mc}{e} = 1.704509024 \cdot10^{-3}\,[\mathrm{T\,m}]. \tag*{(1)}

Consider the scattering of a low energy photon with wavelength λ0\lambda_0 and energy ℏω0\hbar\omega_0 by an electron with energy ε0\varepsilon_0. Denoting the energies of the photon and electron after scattering as ℏω\hbar\omega and ε\varepsilon, their ratio is the dimensionless parameter u=ℏω/εu = \hbar\omega/\varepsilon as in refs. [30, 31].

When γ≫1\gamma\gg1 one can neglect ℏω0≪ε0,ε,ℏω\hbar\omega_0 \ll\varepsilon_0, \varepsilon, \hbar\omega in energy conservation equation and thus

u≡ℏωε=ℏωε0−ℏω=ε0−εε=θeθγ,(2)u \equiv\frac{\hbar\omega}{\varepsilon} = \frac{\hbar\omega}{\varepsilon_0-\hbar\omega} = \frac{\varepsilon_0-\varepsilon}{\varepsilon} = \frac{\theta_e}{\theta_\gamma}, \tag*{(2)}

where θγ\theta_\gamma and θe\theta_e are the scattering angles of photon and electron with respect to the initial electron momentum. The equality εθe=ℏωθγ\varepsilon\theta_e = \hbar\omega\theta_\gamma follows from the law of conservation of the transverse momentum of the system.

Parameter uu lies in the range u∈[0,κ]u \in[0,\kappa], where κ\kappa is twice the initial energy of the photon in the rest frame of the electron, expressed in units of the electron rest energy:

κ=2⋅2γℏω0mc2=4ℏω0ε0(mc2)2.(3)\kappa= 2 \cdot\frac{2\gamma\hbar\omega_0}{mc^2} = \frac{4\hbar\omega_0\varepsilon_0}{(mc^2)^2}. \tag*{(3)}

If α≠π\alpha\ne\pi but tan⁡(α/2)≫1/γ\tan(\alpha/2) \gg1/\gamma, κsin⁡2(α/2)\kappa\sin^2(\alpha/2) should be used instead of κ\kappa. Table 1 shows how the parameters of scattered photon and recoil electron are derived from uu and κ\kappa variables.

scattered photon:recoil electron:
Energy:ℏω=ε0u1+u\hbar\omega= \frac{\varepsilon_0 u}{1+u},ε=ε01+u\varepsilon= \frac{\varepsilon_0}{1+u}.
Edge energy:ℏωmax⁡=ε0κ1+κ\hbar\omega_{\max} = \frac{\varepsilon_0\kappa}{1+\kappa},εmin⁡=ε01+κ\varepsilon_{\min} = \frac{\varepsilon_0}{1+\kappa}.
Scattering angle:θγ=1γκ/u−1\theta_\gamma= \frac{1}{\gamma}\sqrt{\kappa/u-1},θe=uγκ/u−1\theta_e = \frac{u}{\gamma}\sqrt{\kappa/u-1}.
Angle from bend:θ0\theta_0,uθ0u\theta_0.
Maximum bend:θ0\theta_0,Δθ=κθ0\Delta\theta= \kappa\theta_0.
Azimuthal angle:φ\varphi,φ+π\varphi+\pi.
Horizontal angle:ηx=θγcos⁡φ−θ0\eta_x = \theta_\gamma\cos\varphi-\theta_0,θx=uθ0−θecos⁡φ\theta_x = u\theta_0-\theta_e\cos\varphi.
Vertical angle:ηy=θγsin⁡φ\eta_y = \theta_\gamma\sin\varphi,θy=−θesin⁡φ\theta_y=-\theta_e\sin\varphi.

Table 1. Parameters of scattered particles.

Maximum scattering angle of a recoil electron max⁡(θe)=2ℏω0/mc2\max(\theta_e)=2\hbar\omega_0/mc^2 does not depend on the initial electron energy ε0\varepsilon_0 and happens when u=κ/2u=\kappa/2. Since the energy of recoil electron is ε=ε0/(1+u)\varepsilon=\varepsilon_0/(1+u), due to eq. (1) such an electron will be bent by the magnet to the angle θ0(1+u)\theta_0(1+u). According to Figure 1, the first term θ0\theta_0 defines the coordinate system rotation and is denoted in Table 1 as the “angle from bend” and “maximum bend” for scattered photon. The second term uθ0u\theta_0 defines the bend of the recoil electron due to its energy loss. This is the “angle from bend” of recoil electron defined in Table 1. The “maximum bend” of recoil electron Δθ=κθ0\Delta\theta=\kappa\theta_0 happens when u=κu=\kappa and θγ=θe=0\theta_\gamma=\theta_e=0. So far as Δθ/θ0=κ\Delta\theta/\theta_0=\kappa the measurement of the ratio of angles Δθ\Delta\theta to θ0\theta_0 provides the possibility to determine the average energy of electrons in the beam according to the definition of κ\kappa given in eq. (3):

ε0=(mc2)24ℏω0Δθθ0.(4)\varepsilon_0 = \frac{(mc^2)^2}{4\hbar\omega_0}\frac{\Delta\theta}{\theta_0}. \tag*{(4)}

This approach was proposed in ref. [32]. As it follows from eq. (1), Δθ\Delta\theta does not depend on ε0\varepsilon_0:

Δθ=4ℏω0mc2∫B⊥ dlmc/e.(5)\Delta\theta= \frac{4\hbar\omega_0}{mc^2}\frac{\int B_\perp\,dl}{mc/e}. \tag*{(5)}

Since uu is just a fractional part of κ\kappa, the initial energy of an electron ε0\varepsilon_0 does not matter for the “angle from bend” uθ0u\theta_0 of recoil electron with whatever momentum transfer. In a sense, eq. (5) is similar to the formula (1) from ref. [24] which describes the angle of spin θs\theta_s with respect to the direction of motion of a relativistic electron:

θs=μ′μ0∫B⊥dlmc/e,(6)\theta_s = \frac{\mu'}{\mu_0}\frac{\int B_\perp dl}{mc/e}, \tag*{(6)}

where μ′\mu' and μ0\mu_0 are the anomalous and normal parts of the electron magnetic moment.

It is worth noting that there is the established method for the beam energy determination based on direct measurement of the maximal photon energy ℏωmax\hbar\omega_{max} from Table 1, the details could be found in review [33] and references therein.

Cross section

ICS cross section depends on polarization states of all initial and final particles. In case of averaging over the polarizations of the final states the cross section depends solely from the initial photon and electron polarizations. Stokes parameters ξ1,ξ2,ξ3\xi_1,\xi_2,\xi_3 describe the polarization of laser light as it is explained in Appendix § 4. The electron beam polarization has three components ζx,ζy,ζz\zeta_x,\zeta_y,\zeta_z, the total degree of polarization ζx2+ζy2+ζz2∈[0:1]\sqrt{\zeta_x^2+\zeta_y^2+\zeta_z^2}\in[0:1]. We take differential cross section from ref. [34] and after Lorentz transformations it is represented in uu and φ\varphi variables by the sum of the six terms:

1re2dσ0du dφ=1κ(1+u)3[1+(1+u)2−4uκ(1−uκ)(1+u)],1re2dσξ1du dφ=4ξ1κ(1+u)2uκ(1−uκ)cos⁡(2φ),1re2dσξ2du dφ=4ξ2κ(1+u)2uκ(1−uκ)sin⁡(2φ),1re2dσxdu dφ=−2ξ3ζx(1+u)3uκuκ(1−uκ)cos⁡(φ),1re2dσydu dφ=−2ξ3ζy(1+u)3uκuκ(1−uκ)sin⁡(φ),1re2dσzdu dφ=ξ3ζz(1+u)3uκ(u+2)(1−2uκ).(7)\begin{aligned} \frac{1}{r_e^2}\frac{d\sigma_0}{du\,d\varphi} &= \frac{1}{\kappa(1+u)^3}\left[1+(1+u)^2-4\frac{u}{\kappa}\left(1-\frac{u}{\kappa}\right)(1+u)\right],\\ \frac{1}{r_e^2}\frac{d\sigma_{\xi_1}}{du\,d\varphi} &= \frac{4\xi_1}{\kappa(1+u)^2}\frac{u}{\kappa}\left(1-\frac{u}{\kappa}\right)\cos(2\varphi),\\ \frac{1}{r_e^2}\frac{d\sigma_{\xi_2}}{du\,d\varphi} &= \frac{4\xi_2}{\kappa(1+u)^2}\frac{u}{\kappa}\left(1-\frac{u}{\kappa}\right)\sin(2\varphi),\\ \frac{1}{r_e^2}\frac{d\sigma_x}{du\,d\varphi} &= -\frac{2\xi_3\zeta_x}{(1+u)^3}\frac{u}{\kappa}\sqrt{\frac{u}{\kappa}\left(1-\frac{u}{\kappa}\right)}\cos(\varphi),\\ \frac{1}{r_e^2}\frac{d\sigma_y}{du\,d\varphi} &= -\frac{2\xi_3\zeta_y}{(1+u)^3}\frac{u}{\kappa}\sqrt{\frac{u}{\kappa}\left(1-\frac{u}{\kappa}\right)}\sin(\varphi),\\ \frac{1}{r_e^2}\frac{d\sigma_z}{du\,d\varphi} &= \frac{\xi_3\zeta_z}{(1+u)^3}\frac{u}{\kappa}(u+2)\left(1-2\frac{u}{\kappa}\right). \tag*{(7)} \end{aligned}

In (7) rer_e is the classical electron radius and dσ0d\sigma_0 represents the cross section averaged over all polarization terms. Integration of dσ0d\sigma_0 gives the total unpolarized cross section:

σ0=2πre2κ[(1−4κ−8κ2)log⁡(1+κ)+12(1−1(1+κ)2)+8κ].(8)\sigma_0=\frac{2\pi r_e^2}{\kappa}\left[\left(1-\frac{4}{\kappa}-\frac{8}{\kappa^2}\right)\log(1+\kappa)+\frac{1}{2}\left(1-\frac{1}{(1+\kappa)^2}\right)+\frac{8}{\kappa}\right]. \tag*{(8)}

If κ≪1\kappa\ll1 (8) tends to Thomson cross section σ0=83πre2(1−κ)\sigma_0=\frac{8}{3}\pi r_e^2(1-\kappa). The last term in (7) could modify the cross section given by (8) in case ξ3ζz≠0\xi_3\zeta_z\ne0. Its integration gives the result:

σz=ξ3ζz2πre2κ[(1+2κ)log⁡(1+κ)−12(4+(κ1+κ)2)].(9)\sigma_z=\xi_3\zeta_z\frac{2\pi r_e^2}{\kappa}\left[\left(1+\frac{2}{\kappa}\right)\log(1+\kappa)-\frac{1}{2}\left(4+\left(\frac{\kappa}{1+\kappa}\right)^2\right)\right]. \tag*{(9)}

Photons

The cross section with respect to the photon scattering angles ηx\eta_x and ηy\eta_y from Table 1 could be derived from eqs. (7) using the following transformations:

u=κ1+γ2ηx2+γ2ηy2,dudφ=2u2γ2κdηxdηy,(10)u = \frac{\kappa}{1+\gamma^2\eta_x^2+\gamma^2\eta_y^2},\qquad d u d\varphi= \frac{2u^2\gamma^2}{\kappa}d\eta_xd\eta_y, \tag*{(10)}

where we take θ0=0\theta_0=0 for brevity. The parts of the cross section are:

1re2dσ0dηxdηy=2(1+u)3(γuκ)2[1+(1+u)2−4uκ(1−uκ)(1+u)],1re2dσξ1dηxdηy=8ξ1(1+u)2(γuκ)4(ηx2−ηy2),1re2dσξ2dηxdηy=16ξ2(1+u)2(γuκ)4ηxηy,1re2dσxdηxdηy=−4ξ3ζx(1+u)3(γuκ)3uηx,1re2dσydηxdηy=−4ξ3ζy(1+u)3(γuκ)3uηy,1re2dσzdηxdηy=2ξ3ζzγ(1+u)3(γuκ)3(u+2)(κ−2u).(11)\begin{aligned} \frac{1}{r_e^2}\frac{d\sigma_0}{d\eta_xd\eta_y} &= \frac{2}{(1+u)^3}\left(\frac{\gamma u}{\kappa}\right)^2\left[1+(1+u)^2-4\frac{u}{\kappa}\left(1-\frac{u}{\kappa}\right)(1+u)\right],\\ \frac{1}{r_e^2}\frac{d\sigma_{\xi_1}}{d\eta_xd\eta_y} &= \frac{8\xi_1}{(1+u)^2}\left(\frac{\gamma u}{\kappa}\right)^4(\eta_x^2-\eta_y^2),\\ \frac{1}{r_e^2}\frac{d\sigma_{\xi_2}}{d\eta_xd\eta_y} &= \frac{16\xi_2}{(1+u)^2}\left(\frac{\gamma u}{\kappa}\right)^4\eta_x\eta_y,\\ \frac{1}{r_e^2}\frac{d\sigma_x}{d\eta_xd\eta_y} &= \frac{-4\xi_3\zeta_x}{(1+u)^3}\left(\frac{\gamma u}{\kappa}\right)^3u\eta_x,\\ \frac{1}{r_e^2}\frac{d\sigma_y}{d\eta_xd\eta_y} &= \frac{-4\xi_3\zeta_y}{(1+u)^3}\left(\frac{\gamma u}{\kappa}\right)^3u\eta_y,\\ \frac{1}{r_e^2}\frac{d\sigma_z}{d\eta_xd\eta_y} &= \frac{2\xi_3\zeta_z}{\gamma(1+u)^3}\left(\frac{\gamma u}{\kappa}\right)^3(u+2)(\kappa-2u). \tag*{(11)} \end{aligned}

Equations (11) are applicable for describing the spatial distribution of scattered photons measured e. g. by pixel detector with rectangular geometry.

Electrons

To make the formulae in this section more compact and easy to use we introduce modified variables

ϑx=γθx,ϑy=γθy,ϑ0=γθ0.(12)\vartheta_x=\gamma\theta_x,\quad\vartheta_y=\gamma\theta_y,\quad\vartheta_0=\gamma\theta_0. \tag*{(12)}

From definitions given in Table 1 we obtain the equation:

(ϑx−uϑ0)2+ϑy2=u(κ−u),(13)(\vartheta_x-u\vartheta_0)^2+\vartheta_y^2=u(\kappa-u), \tag*{(13)}

whose roots are

u±=κ/2+ϑ0ϑx±κ2/4−ϑx2−ϑy2(1+ϑ02)+κϑ0ϑx1+ϑ02.(14)u_{\pm}=\frac{\kappa/2+\vartheta_0\vartheta_x\pm\sqrt{\kappa^2/4-\vartheta_x^2-\vartheta_y^2(1+\vartheta_0^2)+\kappa\vartheta_0\vartheta_x}}{1+\vartheta_0^2}. \tag*{(14)}

In the ϑx,ϑy\vartheta_x,\vartheta_y plane all recoil electrons are enclosed within the ellipse described by the discriminant of eq. (14). The center of this ellipse is located at the point [ϑx=κϑ0/2;ϑy=0][\vartheta_x=\kappa\vartheta_0/2;\vartheta_y=0], its vertical semi-minor axis equals κ/2\kappa/2 while horizontal semi-major axis is 1+ϑ02\sqrt{1+\vartheta_0^2} times longer. The transformation from u,φu,\varphi to ϑx,ϑy\vartheta_x,\vartheta_y variables is specified by the Jacobian matrix:

\begin{bmatrix} \frac{\partial\vartheta_x}{\partial u} & \frac{\partial\vartheta_x}{\partial\varphi}\\ \frac{\partial\vartheta_y}{\partial u} & \frac{\partial\vartheta_y}{\partial\varphi} \end{bmatrix} = \begin{bmatrix} \vartheta_0-\frac{\kappa/2-u}{\sqrt{u(\kappa-u)}}\cos\varphi& \sqrt{u(\kappa-u)}\sin\varphi\\ -\frac{\kappa/2-u}{\sqrt{u(\kappa-u)}}\sin\varphi& -\sqrt{u(\kappa-u)}\cos\varphi \end{bmatrix}. \tag*{(15)}\end{bmatrix}

The Jacobian determinant:

det⁡(J(u,φ))=κ2−u−ϑ0u(κ−u)cos⁡φ=det⁡(J(ϑx,ϑy))=±κ24−ϑx2−ϑy2(1+ϑ02)+κϑ0ϑx.(16)\begin{aligned} \det(J(u,\varphi)) = \frac{\kappa}{2} - u - \vartheta_0\sqrt{u(\kappa-u)}\cos\varphi\\ &= \det(J(\vartheta_x,\vartheta_y)) = \pm\sqrt{\frac{\kappa^2}{4}-\vartheta_x^2-\vartheta_y^2(1+\vartheta_0^2)+\kappa\vartheta_0\vartheta_x}. \tag*{(16)} \end{aligned}

D=∣det⁡(J(ϑx,ϑy))∣D=\lvert\det(J(\vartheta_x,\vartheta_y))\rvert equals the absolute value of the square root of discriminant in eq. (14). The following expressions follow from the definitions given in Table 1:

cos⁡(φ)=uϑ0−ϑxu(κ−u),sin⁡(φ)=−ϑyu(κ−u),\cos(\varphi)=\frac{u\vartheta_0-\vartheta_x}{\sqrt{u(\kappa-u)}},\qquad \sin(\varphi)=\frac{-\vartheta_y}{\sqrt{u(\kappa-u)}},
cos⁡(2φ)=(uϑ0−ϑx)2−ϑy2u(κ−u),sin⁡(2φ)=−2(uϑ0−ϑx)ϑyu(κ−u).(17)\cos(2\varphi)=\frac{(u\vartheta_0-\vartheta_x)^2-\vartheta_y^2}{u(\kappa-u)},\qquad \sin(2\varphi)=\frac{-2(u\vartheta_0-\vartheta_x)\vartheta_y}{u(\kappa-u)}. \tag*{(17)}

Lets apply the variables transformation to eqs. (7):

1re2dσ0dϑxdϑy=1Dκ(1+u)3[1+(1+u)2−4u(κ−u)κ2(1+u)],1re2dσξ1dϑxdϑy=4ξ1Dκ3(1+u)2((uϑ0−ϑx)2−ϑy2),1re2dσξ2dϑxdϑy=−8ξ2Dκ3(1+u)2(uϑ0−ϑx)ϑy,1re2dσxdϑxdϑy=−2ξ3ζxDκ2(1+u)3u(uϑ0−ϑx),1re2dσydϑxdϑy=2ξ3ζyDκ2(1+u)3uϑy,1re2dσzdϑxdϑy=ξ3ζzDκ2(1+u)3u(u+2)(κ−2u).(18)\begin{aligned} \frac{1}{r_e^2}\frac{d\sigma_0}{d\vartheta_xd\vartheta_y} &=\frac{1}{D\kappa(1+u)^3}\left[1+(1+u)^2-4\frac{u(\kappa-u)}{\kappa^2}(1+u)\right],\\ \frac{1}{r_e^2}\frac{d\sigma_{\xi_1}}{d\vartheta_xd\vartheta_y} &=\frac{4\xi_1}{D\kappa^3(1+u)^2}\left((u\vartheta_0-\vartheta_x)^2-\vartheta_y^2\right),\\ \frac{1}{r_e^2}\frac{d\sigma_{\xi_2}}{d\vartheta_xd\vartheta_y} &=\frac{-8\xi_2}{D\kappa^3(1+u)^2}(u\vartheta_0-\vartheta_x)\vartheta_y,\\ \frac{1}{r_e^2}\frac{d\sigma_x}{d\vartheta_xd\vartheta_y} &=\frac{-2\xi_3\zeta_x}{D\kappa^2(1+u)^3}u(u\vartheta_0-\vartheta_x),\\ \frac{1}{r_e^2}\frac{d\sigma_y}{d\vartheta_xd\vartheta_y} &=\frac{2\xi_3\zeta_y}{D\kappa^2(1+u)^3}u\vartheta_y,\\ \frac{1}{r_e^2}\frac{d\sigma_z}{d\vartheta_xd\vartheta_y} &=\frac{\xi_3\zeta_z}{D\kappa^2(1+u)^3}u(u+2)(\kappa-2u). \tag*{(18)} \end{aligned}

The total cross section is defined by the sum of these terms, each of which is calculated twice (for u+u_+ and for u−u_-). With the following linear transformation instead of the ellipse we obtain the circle of unit radius with the center located at the origin of new coordinate system:

x=Aϑxκ/2−B;y=ϑyκ/2,wherex=A\frac{\vartheta_x}{\kappa/2}-B;\qquad y=\frac{\vartheta_y}{\kappa/2},\quad\text{where}
A=11+ϑ02andB=ϑ01+ϑ02.(19)A=\frac{1}{\sqrt{1+\vartheta_0^2}}\quad\text{and}\quad B=\frac{\vartheta_0}{\sqrt{1+\vartheta_0^2}}. \tag*{(19)}

The condition ϑ0≫1\vartheta_0\gg1 obviously provides better separation of recoil electrons from the beam and in this case A≃1/ϑ0A\simeq1/\vartheta_0 and B≃1B\simeq1. Exact calculations give the following:

dϑxdϑyD=κ dx dy21−x2−y2,u±=κ2(1+Bx±A1−x2−y2)=κ2(1+Δ±),(uϑ0−ϑx)±=κ2(−Ax±B1−x2−y2)=κ2δ±.(20)\begin{aligned} \frac{d\vartheta_xd\vartheta_y}{D}&=\frac{\kappa\,dx\,dy}{2\sqrt{1-x^2-y^2}},\\ u_\pm&=\frac{\kappa}{2}\left(1+Bx\pm A\sqrt{1-x^2-y^2}\right)=\frac{\kappa}{2}(1+\Delta_\pm),\\ (u\vartheta_0-\vartheta_x)_\pm&=\frac{\kappa}{2}\left(-Ax\pm B\sqrt{1-x^2-y^2}\right)=\frac{\kappa}{2}\delta_\pm. \tag*{(20)} \end{aligned}

Parameters Δ±\Delta^{\pm} and δ±\delta^{\pm} in eqs. (20) are the shorthand’s for corresponding parenthesized expressions. We see that dϑxdϑy/Dd\vartheta_xd\vartheta_y/D provides the cross section “amplification” near the border of the circle, while the difference u+−u−u_{+}-u_{-} is at maximum in the center of the circle. Finally we obtain:

1re2dσ0dx dy∣±=1+(1+u±)2−(1+u±)(1−Δ±2)2(1+u±)31−x2−y2,1re2dσξ1dx dy∣±=ξ1(δ±2−y2)2(1+u±)21−x2−y2,1re2dσξ2dx dy∣±=−ξ2δ±y(1+u±)21−x2−y2,1re2dσxdx dy∣±=ξ3ζx(−u±δ±)2(1+u±)31−x2−y2,1re2dσydx dy∣±=ξ3ζyu±y2(1+u±)31−x2−y2,1re2dσzdx dy∣±=ξ3ζz[−u±(u±+2)Δ±]2(1+u±)31−x2−y2.(21)\begin{aligned} \left.\frac{1}{r_e^2}\frac{d\sigma_0}{dx\,dy}\right|_{\pm} &= \frac{1+(1+u_{\pm})^2-(1+u_{\pm})(1-\Delta_{\pm}^2)}{2(1+u_{\pm})^3\sqrt{1-x^2-y^2}},\\ \left.\frac{1}{r_e^2}\frac{d\sigma_{\xi_1}}{dx\,dy}\right|_{\pm} &= \frac{\xi_1(\delta_{\pm}^2-y^2)}{2(1+u_{\pm})^2\sqrt{1-x^2-y^2}},\\ \left.\frac{1}{r_e^2}\frac{d\sigma_{\xi_2}}{dx\,dy}\right|_{\pm} &= \frac{-\xi_2\delta_{\pm}y}{(1+u_{\pm})^2\sqrt{1-x^2-y^2}},\\ \left.\frac{1}{r_e^2}\frac{d\sigma_x}{dx\,dy}\right|_{\pm} &= \frac{\xi_3\zeta_x(-u_{\pm}\delta_{\pm})}{2(1+u_{\pm})^3\sqrt{1-x^2-y^2}},\\ \left.\frac{1}{r_e^2}\frac{d\sigma_y}{dx\,dy}\right|_{\pm} &= \frac{\xi_3\zeta_yu_{\pm}y}{2(1+u_{\pm})^3\sqrt{1-x^2-y^2}},\\ \left.\frac{1}{r_e^2}\frac{d\sigma_z}{dx\,dy}\right|_{\pm} &= \frac{\xi_3\zeta_z[-u_{\pm}(u_{\pm}+2)\Delta_{\pm}]}{2(1+u_{\pm})^3\sqrt{1-x^2-y^2}}. \tag*{(21)} \end{aligned}

The cross section in x,yx,y plane is the sum of “+” and “-” solutions. It’s time to explain why the variables xx and yy were introduced. To build a function that describes the number of events in a pixel of the detector of recoil electrons, we need the integral of the cross section over a rectangular area. The antiderivative of f(x,y)=1/1−x2−y2f(x,y)=1/\sqrt{1-x^2-y^2} function is an analytical expression

F(x,y)=xatan2⁡(y,ρ)+yatan2⁡(x,ρ)−atan2⁡(xy,ρ),(22)F(x,y)=x\operatorname{atan2}(y,\rho)+y\operatorname{atan2}(x,\rho)-\operatorname{atan2}(xy,\rho), \tag*{(22)}

where ρ=1−x2−y2\rho=\sqrt{1-x^2-y^2}. The integral over the pixel with the corners x0,y0x_0,y_0 and x1,y1x_1,y_1 is

I(x0,y0,x1,y1)=F(x0,y0)+F(x1,y1)−F(x0,y1)−F(x1,y0).(23)I(x_0,y_0,x_1,y_1)=F(x_0,y_0)+F(x_1,y_1)-F(x_0,y_1)-F(x_1,y_0). \tag*{(23)}

Monte Carlo experiment

General considerations

Consider again the diagram shown in Figure 1. In the Monte-Carlo experiment scattered photons and electrons hit the notional detectors located far enough from the interaction area and the bending magnet. These detectors are installed along the xx-axis on the right-side of Figure 1. It is also worth paying attention that there are no magnetic elements between the magnet and the detectors.

If R0R_0 is the bending radius of a beam electron, then the radius of the recoil electron with minimal energy equals Rmin⁡=R0/(1+κ)R_{\min}=R_0/(1+\kappa). Consider the bending dipole of length LL with rectangular shape of the magnetic poles. All the electrons enter the dipole along the same line – the beam orbit. The equality Δθ=κθ0\Delta\theta=\kappa\theta_0 in Table 1 was obtained under the assumption that the field integral does not depend on the electron energy. The accuracy of this assumption depends on the parameters combination Λ=κL/2R0\Lambda=\kappa L/2R_0. At the end of the dipole an electron with minimal energy εmin⁡\varepsilon_{\min} will have transverse horizontal distance from the beam orbit Δx≈ΛL\Delta x\approx\Lambda L, which sets the requirement for a transverse size of the uniform field region inside the magnet. Relative difference in the arcs lengths inside the magnet for ε0\varepsilon_0 and εmin⁡\varepsilon_{\min} electrons is estimated as ΔL/L≃Λ2/2\Delta L/L \simeq\Lambda^2/2. With Λ≃10−3\Lambda\simeq10^{-3} the relative difference in the magnetic field integrals ΔL/L≃5⋅10−7\Delta L/L \simeq5 \cdot10^{-7}. Small enough values of Λ\Lambda are accessible with small bending angles L/R0≪1L/R_0 \ll1 when κ≥1\kappa\ge1. At low electron beam energies, when κ≪1\kappa\ll1, the bending angle could be increased.

Use of two-dimensional pixel detector for recoil electrons was proposed for transverse beam polarization measurement in ref. [35]. Figure 2 illustrates the scattered photons, recoil electrons and corresponding detectors in x,yx,y plane.

Image not extracted

Figure 2. Detection plane. 2D pixel detectors for photons and electrons are represented by dotted rectangles.

It is assumed that the detectors are precisely positioned relative to each other. Each detector has a certain geometry, number of pixels and pixel size in both dimensions. In the numerical experiment the detectors are represented by the two-dimensional (x,y)(x,y) histograms where each bin corresponds to one pixel. The following parameters are used in further Monte-Carlo simulations:

  • γ\gamma is the mean value of the beam electron Lorentz factor.

  • θ0\theta_0 is the equilibrium orbit bending angle.

  • λ0\lambda_0 and ℏω0=hc/λ0\hbar\omega_0 = hc/\lambda_0 are the wavelength and energy of a laser photon.

  • κ=4γℏω0/mc2\kappa= 4\gamma\hbar\omega_0/mc^2 is the mean value of the Compton scattering parameter κ\kappa.

  • Stokes vector [ξ1,ξ2,ξ3][\xi_1,\xi_2,\xi_3] describes the laser beam polarization.

  • Vector [ζx,ζy,ζz][\zeta_x,\zeta_y,\zeta_z] describes the electron beam polarization.

  • uu and φ\varphi parameters are generated according to the sum of eqs. (7).

  • Lorentz factor of a beam electron is γ′=γ(1+Δ)\gamma' = \gamma(1 + \Delta) where Δ\Delta is extracted from the normal distribution with zero mean value and standard deviation described by the relative energy spread of the electron beam σγ/γ\sigma_\gamma/\gamma.

  • The transverse horizontal coordinate of the electron at the moment of scattering due to synchrotron motion is DxΔD_x\Delta where DxD_x is the dispersion function.

  • The transverse coordinates xx, yy of the electron due to betatron motion are generated by the normal distributions with zero mean value and σx=ϵxβx\sigma_x = \sqrt{\epsilon_x\beta_x}, σy=ϵyβy\sigma_y = \sqrt{\epsilon_y\beta_y} described by corresponding emittance and β\beta-function.

  • The transverse angles x′x', y′y' of the electron due to betatron motion are generated by the normal distributions with zero mean value and σx′=ϵx/βx\sigma'_x = \sqrt{\epsilon_x/\beta_x}, σy′=ϵy/βy\sigma'_y = \sqrt{\epsilon_y/\beta_y}.

  • The length of the interaction region of laser radiation with electron beam is assumed to be very short (≪L1\ll L_1) and thus it is not taken into account.

  • The XγX_\gamma, YγY_\gamma, XeX_e, YeY_e coordinates of scattered photon and electron at the detection plane are obtained by the following equations:

Xγ=1cos⁡θ0(DxΔ+x+L1tan⁡[x′+cos⁡φκ/u−1γ(1+Δ)])−L2tan⁡θ0,Yγ=y+L1tan⁡[y′+sin⁡φκ/u−1γ(1+Δ)],Xe=1cos⁡θ0(DxΔ+x+L1tan⁡[x′−cos⁡φu(κ−u)γ])+L2tan⁡θ0[u−Δ1+Δ],Ye=y+L1tan⁡[y′−sin⁡φu(κ−u)γ].(24)\begin{aligned} X_\gamma&= \frac{1}{\cos\theta_0}\left(D_x\Delta+x+L_1\tan\left[x'+\frac{\cos\varphi\sqrt{\kappa/u-1}}{\gamma(1+\Delta)}\right]\right)-L_2\tan\theta_0,\\ Y_\gamma&= y+L_1\tan\left[y'+\frac{\sin\varphi\sqrt{\kappa/u-1}}{\gamma(1+\Delta)}\right],\\ X_e &= \frac{1}{\cos\theta_0}\left(D_x\Delta+x+L_1\tan\left[x'-\frac{\cos\varphi\sqrt{u(\kappa-u)}}{\gamma}\right]\right)+L_2\tan\theta_0\left[u-\frac{\Delta}{1+\Delta}\right],\\ Y_e &= y+L_1\tan\left[y'-\frac{\sin\varphi\sqrt{u(\kappa-u)}}{\gamma}\right]. \tag*{(24)} \end{aligned}

If the photon XγX_\gamma, YγY_\gamma or electron XeX_e, YeY_e coordinates hits the pixel of the photon/electron detector, the content of the corresponding histogram bin is increased by one. Note that the eqs. (24) contain the lattice functions βx\beta_x, βy\beta_y, DxD_x but not their derivatives. Each beam electron at the area of interaction with laser radiation has some deviations from the equilibrium parameters due to betatron and synchrotron motion. The transverse coordinates of this electron at the plane of the detectors installation could be obtained from the full set of its initial parameters by tracking through the spectrometer. According to eq. (5) bending of the recoil electrons differs from bending of the beam electrons by an additional angle which does not depend on the initial electron parameters. It means that the (gaussian) blur of the two-dimensional distribution of recoil electrons described by eqs. (21) is equal to the transverse beam size (σx\sigma_x, σy\sigma_y) at the plane of detectors.

It is worth noting that in this simulation it is implicitly assumed that the scattering probability does not depend on the transverse coordinates of the electron. This condition is satisfied if the size of the laser spot significantly exceeds the size of the electron beam. Such a configuration of the interaction area is necessary in any case when we are going to measure the true average parameters of the electron beam.

FCC-ee beam polarimeter

FCC-ee is a lepton collider with centre-of-mass energies between 90 and 350 GeV [36]. It is considered as a potential intermediate step towards the realization of the hadron facility. Beam energy calibration by resonant depolarization is the basis for the precise measurements of the ZZ mass and width with a precision of <100<100 keV, and of the WW mass and width with a precision of the order of 500 keV. The conceptual design of the Compton polarimeter may be found in [36, 37] and the same setup is considered here. Table 2 contains the list of simulation parameters which will be

ε0=45.6\varepsilon_0 = 45.6 GeVγ=89237\gamma= 89237ϵx=270\epsilon_x = 270 pm radβx=100\beta_x = 100 mL1=117L_1 = 117 m
λ0=532\lambda_0 = 532 nmκ=1.6279\kappa= 1.6279ϵy=1\epsilon_y = 1 pm radβy=20\beta_y = 20 mL2=100L_2 = 100 m
ω0=2.331\omega_0 = 2.331 eVϑ0=190.44\vartheta_0 = 190.44θ0=2.1341\theta_0 = 2.1341 mradDx=25D_x = 25 mmσγ/γ=0.001\sigma_\gamma/\gamma= 0.001

Table 2. Simulation parameters. All designations have been defined in the text above.

identical in all numerical experiments below. Bunch revolution frequency at FCC-ee is 3⋅103 s−13 \cdot10^3\ \mathrm{s}^{-1} while the rate of Compton scattering events is estimated as 2⋅106 s−12 \cdot10^6\ \mathrm{s}^{-1}. The parameters of the pixel detectors for photons and electrons are presented in Table 3.

DetectorSize (X×YX \times Y)NpixN_{\mathrm{pix}} (X×YX \times Y)Pixel size (X×YX \times Y)
Photons10×1010 \times10 mm100×100100 \times100100×100 μm100 \times100\ \mu\mathrm{m}
Electrons400×4400 \times4 mm1600×801600 \times80250×50 μm250 \times50\ \mu\mathrm{m}

Table 3. Detectors: geometry, number of pixels, size of pixels.

Experiment I

The goal of the first numerical experiment is to verify the above formulae for the cross sections and check the fitting procedure. The subject of this study is the trivial case when the Stokes vector of laser polarization is [ξ1,ξ2,ξ3]=[0,0,1][\xi_1,\xi_2,\xi_3] = [0,0,1] and the electron beam is unpolarized ζx,ζy,ζz=[0,0,0]\zeta_x,\zeta_y,\zeta_z = [0,0,0]. Figure 3 shows the distribution of scattered photons at the detector obtained with 10710^7 Monte-Carlo

Image not extracted

Figure 3. Photon spot at the detector. Number of bins is reduced for better visualization.

events. In FCC-ee conditions [37] this amount is available within 5 s measurement time with approximately 1.5⋅1041.5 \cdot10^4 collisions between laser pulse and electron bunch.

The fit function for the scattered photons distribution is based on eqs. (11) with nine fit parameters X0X_0, Y0Y_0, σx\sigma_x, σy\sigma_y, ξ1\xi_1, ξ2\xi_2, ξ3ζx\xi_3\zeta_x, ξ3ζy\xi_3\zeta_y, ξ3ζz\xi_3\zeta_z representing the spot center position, blur effect from electron beam emittance and five independent polarization components. The tenth parameter is the normalization. The distance from the interaction area to the detector L1L_1 and the scattering parameter κ\kappa are fixed according to the setup. The results of the fit are presented in Table 4 (NDF is the acronym for Number of Degrees of Freedom).

X0=−213.538±0.001X_0=-213.538 \pm0.001 mmY0=−0.002±0.001Y_0=-0.002 \pm0.001 mm
σx=255±3\sigma_x=255 \pm3 μ\mumσy=30±18\sigma_y=30 \pm18 μ\mum
ξ1=0.000±0.002\xi_1=0.000 \pm0.002ξ2=−0.001±0.001\xi_2=-0.001 \pm0.001
ξ3ζx=0.004±0.006\xi_3\zeta_x=0.004 \pm0.006ξ3ζy=−0.008±0.006\xi_3\zeta_y=-0.008 \pm0.006
ξ3ζz=0.000±0.002\xi_3\zeta_z=0.000 \pm0.002χ2/NDF=9796.9/9990\chi^2/\mathrm{NDF}=9796.9/9990

Table 4. Photon spot fit results.

It can be seen that the position of the spot is established with accuracy about 11 μ\mum. The σx\sigma_x and σy\sigma_y values carry the information about the combination of beam parameters at the interaction area. The accuracy of σy\sigma_y is poor because it is three times smaller than the pixel size. It would be useful to recall that in order to detect gamma rays, they will be converted into e+e−e^+e^- pairs, which will lead to additional blurring and reduce the number of registered events. However all polarization parameters correspond to their zero set values with absolute accuracy of better than one percent.

Let us take advantage of the opportunities provided by the numerical experiment and place the electron detector so that it detects all recoil electrons, including those that propagate very close to the beam. Such a distribution of scattered electrons is presented in Figure 4.

Image not extracted

Figure 4. Recoil electrons at the detector. Number of xx-bins is reduced to 100 for better visualization. Electrons detector covers the beam at the coordinates X=0X=0, Y=0Y=0, which is impossible to do in practice.

The fit procedure is done by Pearson’s method, the expected bin error is estimated as 1+F\sqrt{1+\mathcal{F}}, where the fit function F(X,Y)\mathcal{F}(X,Y) is calculated in three steps. First, geometry parameters X1X_1, X2X_2, Y1Y_1, Y2Y_2 shown in Figure 2 specify ellipse of scattered electrons defined in eqs. (14) and (16). With linear transformations, based on the known detector geometry, these parameters are converted to dimensionless variables xx and yy defined in eqs. (19). The integrals from (23) are calculated as well as the corresponding cross section terms from eqs. (21) at the center of each pixel. Scattering parameter κ\kappa is fixed as it was done with photons while neither L1L_1 nor L2L_2 are not used. Finally the distribution, obtained by pixel by pixel multiplication of the integrals and the cross sections, is convoluted with two-dimensional Gaussian to account for transverse beam sizes at the detection plane. Totally the fit function has twelve parameters including normalization, while only non-empty bins are accounted in NDF. The results of the fit are presented in Table 5.

X1=0.0035±0.0016X_1 = 0.0035 \pm0.0016 mmX2=347.635±0.003X_2 = 347.635 \pm0.003 mm
Y1=−1.0682±0.0001Y_1 = -1.0682 \pm0.0001 mmY2=1.0682±0.0001Y_2 = 1.0682 \pm0.0001 mm
σx=320.7±1.5\sigma_x = 320.7 \pm1.5 μ\mumσy=27.06±0.03\sigma_y = 27.06 \pm0.03 μ\mum
ξ1=0.001±0.001\xi_1 = 0.001 \pm0.001ξ2=0.432±0.198\xi_2 = 0.432 \pm0.198
ξ3ζx=1.000±0.195\xi_3\zeta_x = 1.000 \pm0.195ξ3ζy=−0.001±0.002\xi_3\zeta_y = -0.001 \pm0.002
ξ3ζz=0.000±0.001\xi_3\zeta_z = 0.000 \pm0.001χ2/NDF=51568.9/52270\chi^2/\mathrm{NDF} = 51568.9/52270
ε0=45.5997±0.0007\varepsilon_0 = 45.5997 \pm0.0007 GeV

Table 5. Scattered electrons ellipse fit results.

It can be seen that the accuracy of the left ellipse border position X1X_1 is 0.160.16 μ\mum which is approximately twice better than the accuracy of the right border position X2X_2. This is explained by larger amount of recoil electrons near X1X_1 (the beam position). The difference Y2−Y1Y_2-Y_1 is defined with relative precision of 5⋅10−55 \cdot10^{-5}, which allows to measure (or just check) L1L_1 with the same accuracy. The X1X_1 and (Y1+Y2)/2(Y_1+Y_2)/2 define the electron beam transverse position relative to the detector. Unlike the photons distribution, the σx\sigma_x and σy\sigma_y here are the parameters describing the transverse beam sizes at the detector position and we see that both sizes could be measured with rather high accuracy (compared to e. g. ref. [38]). As for polarization parameters, ξ2\xi_2 and ξ3ζx\xi_3\zeta_x determination is wrong. This happens because the sum of u+u_+ and u−u_- solutions in corresponding terms of the cross section (21) have opposite signs and almost cancel each other when ϑ0≫1\vartheta_0 \gg1 (in our case ϑ0=γθ0≃190\vartheta_0=\gamma\theta_0 \simeq190). The other three polarization parameters ξ1\xi_1, ξ3ζy\xi_3\zeta_y and ξ3ζz\xi_3\zeta_z however are determined even more precise than from the photons distribution. The value of the beam energy ε0\varepsilon_0 in the last row of Table 5 is obtained from X0X_0, X1X_1, X2X_2 fit parameters with relative accuracy of about 1.5⋅10−51.5 \cdot10^{-5} by the formula which follows from eq. (5):

ε0=(mc2)24ω0X2−X1X1−X0.(25)\varepsilon_0 = \frac{(mc^2)^2}{4\omega_0}\frac{X_2-X_1}{X_1-X_0}. \tag*{(25)}

Experiment II

In this experiment we study more realistic case when the recoil electron detector is shifted 15 mm away from the electron beam. It is assumed that this gap will provide a sufficient physical aperture for the electron beam. The Stokes vector of laser polarization is chosen as [ξ1,ξ2,ξ3]=[0.1,0.1,0.99][\xi_1,\xi_2,\xi_3] = [0.1,0.1,0.99] with small amount of vestigial linear polarization which may exist due to imperfect polarization control. The set of the electron beam polarization parameters is [ζx,ζy,ζz]=[0.1,0.25,0.1][\zeta_x,\zeta_y,\zeta_z] = [0.1,0.25,0.1]. The expected polarization of the pilot electron bunches at FCC-ee (averaged over thousands beam revolutions) has ζy\zeta_y component only. The ζx\zeta_x and ζz\zeta_z components are added to the simulations in order to investigate the possibility of measuring the electron beam polarization in general case.

We do not present the figure with the distribution of photons here since visually it is very close to that shown in Figure 3. The fitting procedure is the same as in previous experiment and the results obtained from the fit are presented in Table 6. From the table we can conclude that all polarization parameters are determined correctly with absolute accuracies from 0.1% to 0.7%.

X0=−213.539±0.002 mmX_0 = -213.539 \pm0.002\ \mathrm{mm}Y0=0.000±0.001 mmY_0 = 0.000 \pm0.001\ \mathrm{mm}
σx=246±4 μm\sigma_x = 246 \pm4\ \mu\mathrm{m}σy=13±70 μm\sigma_y = 13 \pm70\ \mu\mathrm{m}
ξ1=0.102±0.002\xi_1 = 0.102 \pm0.002ξ2=0.100±0.001\xi_2 = 0.100 \pm0.001
ξ3ζx=0.095±0.007\xi_3\zeta_x = 0.095 \pm0.007ξ3ζy=0.247±0.006\xi_3\zeta_y = 0.247 \pm0.006
ξ3ζz=0.105±0.002\xi_3\zeta_z = 0.105 \pm0.002χ2/NDF=9935.8/9990\chi^2/\mathrm{NDF} = 9935.8/9990

Table 6. Photon spot fit results.

The distribution of recoil electrons in Figure 5 looks different mainly due to the new location of the detector, the results of the fit are presented in Table 7.

Image not extracted

Figure 5. Recoil electrons at the detector. Number of xx-bins is reduced to 100 for better visualization. Electrons detector is shifted in XX by 15 mm away from the beam coordinates X=0X = 0, Y=0Y = 0.

X1=0.013±0.007 mmX_1 = 0.013 \pm0.007\ \mathrm{mm}X2=347.632±0.004 mmX_2 = 347.632 \pm0.004\ \mathrm{mm}
Y1=−1.0682±0.0001 mmY_1 = -1.0682 \pm0.0001\ \mathrm{mm}Y2=1.0684±0.0001 mmY_2 = 1.0684 \pm0.0001\ \mathrm{mm}
σx=319.6±4.3 μm\sigma_x = 319.6 \pm4.3\ \mu\mathrm{m}σy=27.15±0.03 μm\sigma_y = 27.15 \pm0.03\ \mu\mathrm{m}
ξ1=0.100±0.001\xi_1 = 0.100 \pm0.001ξ2=0.100\xi_2 = 0.100
ξ3ζx=0.099\xi_3\zeta_x = 0.099ξ3ζy=0.246±0.002\xi_3\zeta_y = 0.246 \pm0.002
ξ3ζz=0.099±0.001\xi_3\zeta_z = 0.099 \pm0.001χ2/NDF=50152.7/51245\chi^2/\mathrm{NDF} = 50152.7/51245
ε0=45.5959±0.0025 GeV\varepsilon_0 = 45.5959 \pm0.0025\ \mathrm{GeV}

Table 7. Scattered electrons ellipse fit results.

The fitting was performed in the same way as in our previous experiment, except that the parameters ξ2\xi_2 and ξ3ζx\xi_3\zeta_x are fixed according to the results obtained from the photons distribution. We see that the accuracy of determining parameters X1X_1 and X2X_2 is almost the same now. The transverse beam sizes σx\sigma_x and σy\sigma_y are in perfect agreement with the results of the first experiment. The polarization parameters ξ1\xi_1, ξ3ζy\xi_3\zeta_y and ξ3ζz\xi_3\zeta_z are determined with absolute accuracy of 0.1%, 0.2% and 0.1%. The beam energy ε0\varepsilon_0 is “measured” with relative accuracy of about 5.5⋅10−55.5 \cdot10^{-5} and differs from the true value by a little less than two standard deviations. The accuracy is worse than in the first experiment cause the electrons detector is shifted away from the electron beam by 15 mm and X1X_1 parameter is now determined by extrapolation of the shape of distribution.

Summary

Simultaneous measurement of the spatial distributions of gamma quanta and recoil electrons makes it possible to significantly expand the capabilities of the Compton polarimeter. In addition to all polarization components of the laser and electron beams, another parameters can be measured with rather high accuracy: the average energy of beam electrons, the integral of the field in the spectrometer magnet, the transverse coordinates and dimensions of the electron beam at the location of the detectors. All these parameters are obtained by analysis of the two-dimensional spatial distribution of recoil electrons. The recoil electrons have a sharp boundary in the scattering angle which, after bending in the magnet, takes the form of an ellipse in the detection plane. The sharp edges of the ellipse are blurred by the gaussian beam sizes of the electron beam at the location of the detectors.

The considered approach is applicable when the vertical size of the ellipse of recoil electrons is larger than the vertical beam size:

L12ω0mc2>εyβy.(26)L_1 \frac{2\omega_0}{mc^2} > \sqrt{\varepsilon_y\beta_y}. \tag*{(26)}

Recoil electrons propagate to the inner side of the beam orbit and thus there is no direct background from high energy synchrotron radiation. The angular distribution of scattered photons partially overlaps with photons of synchrotron radiation, this circumstance requires additional consideration if we are dealing with high electron energies. Electrons are registered directly by ionization losses while photons should be converted into pairs leading to a decrease in the detection efficiency and spatial resolution. Despite the fact that the fluxes of scattered photons and electrons are almost the same, the flux density of electrons is much lower due to bending and corresponding spatial separation by energies, making easier the simultaneous detection of multiple recoil electrons.

The main technical difficulty for the practical implementation of the proposed measurement scheme is the need to have a long spectrometer arm, in which the recoil electrons are not exposed to magnetic fields. A high degree of field uniformity in the bending magnet is required for accurate determination of the beam energy. Additional simulations with realistic parameters of the magnetic system should be carried out to estimate the systematic effects due to these factors. If the realistic situation differs from the ideal one, the successful fitting of the recoil electron distributions can be performed by introducing additional fit parameters to the transformation from the detector coordinates XX and YY ( Figure 2) to the dimensionless coordinates xx and yy used in eqs. (21).

The FCC-ee polarimeter was designed to provide fast and precise measurements of the transverse polarization of the pilot bunches. This will allow to continuously monitor the beam energy by resonant depolarization technique at the beam energies up to 80 GeV approximately, when it becomes impossible to keep the beam polarized. For these needs the analysis of the distribution of recoil electrons will provide more accurate measurements at the same acquisition time if compared with scattered photons. Meanwhile the energy spectrometer option of the polarimeter could be cross-checked with the resonant depolarization results and allow to use it at higher beam energies and for interpolation purposes. Detection and analysis of distributions of both photons and recoil electrons will reduce systematic errors in measurements of all polarization components. Accurate determination of the transverse sizes of the electron beam is, in a sense, a usefull additional bonus.

Stokes parameters

The Stokes parameters describe the polarization state of electromagnetic radiation. Their definition is slightly different in different sources, so below are the definition that we use here.

  • ξ0=Ex2+Ey2\xi_0 = E_x^2 + E_y^2 is the intensity of light. With normalization Ex2+Ey2=1E_x^2 + E_y^2 = 1 for 100% polarized laser radiation ξ0=ξ12+ξ22+ξ32=1\xi_0 = \sqrt{\xi_1^2 + \xi_2^2 + \xi_3^2} = 1.

  • ξ1=Ex2−Ey2\xi_1 = E_x^2 - E_y^2.

    Ex=1,Ey=0,ξ1=+1E_x = 1, E_y = 0, \xi_1 = +1: 100% linear polarization along x-axis.

    Ex=0,Ey=1,ξ1=−1E_x = 0, E_y = 1, \xi_1 = -1: 100% linear polarization along y-axis.

  • ξ2=2ExEycos⁡(δ)\xi_2 = 2E_xE_y\cos(\delta).

    Ex=Ey,δ=0,ξ2=+1E_x = E_y, \delta= 0, \xi_2 = +1: 100% linear polarization along φ=+π/4\varphi= +\pi/4.

    Ex=Ey,δ=π,ξ2=−1E_x = E_y, \delta= \pi, \xi_2 = -1: 100% linear polarization along φ=−π/4\varphi= -\pi/4.

  • ξ3=2ExEysin⁡(δ)\xi_3 = 2E_xE_y\sin(\delta).

    Ex=Ey,δ=+π/2,ξ3=+1E_x = E_y, \delta= +\pi/2, \xi_3 = +1: 100% right circular polarization.

    Ex=Ey,δ=−π/2,ξ3=−1E_x = E_y, \delta= -\pi/2, \xi_3 = -1: 100% left circular polarization.

Acknowledgments

The author expresses his gratitude to Mikhail Achasov and Ivan Koop for useful discussions, to Alain Blondel and Eugene Levichev for their interest and motivation to continue these studies. The author would also like to thank Katsunobu Oide who always found a solution to the difficult issue of polarimeter integration to the FCC-ee lattice (otherwise these studies would have been closed in the early stages of discussion) and Aurélian Martens for his help in deriving the cross section formulae. This study is dedicated to the upcoming centennial anniversary of the discovery of the Compton effect.

References

  1. [1]R.H. Milburn, Electron Scattering by an Intense Polarized Photon Field, Phys. Rev. Lett. 10 (1963) 75.DOI
  2. [2]F.R. Arutyunian and V.A. Tumanian, The Compton effect on relativistic electrons and the possibility of obtaining high energy beams, Phys. Lett. 4 (1963) 176.
  3. [3]J. Ballam, G.B. Chadwick, R. Gearhart, Z.G.T. Guirgossian, P.R. Klein, A. Levy et al., Total and Partial Photoproduction Cross Sections at 1.44, 2.8, and 4.7 GeV, Phys. Rev. Lett. 23 (1969) 498.
  4. [4]A.A. Sokolov and M. Ternov, On polarization and spin effects in the theory of synchrotron radiation, Sov. Phys. - Dokl. 8 (1964) 1203.
  5. [5]Belbeoch, R., Berger, M., Le-Duc, J., Level, M. P., Marin, P., Nguyen Ngoc, H. et al., New experiments with ACO, in Nat. Part. Acc. Conf., vol. 681009, (Moscow, USSR), pp. 129 – 133, 1968.
  6. [6]Y.S. Derbenev, A.M. Kondratenko, S.I. Serednyakov, A.N. Skrinsky, G.M. Tumaikin and Y.M. Shatunov, Radiative Polarization: Obtaining, Control, Using, Part. Accel. 8 (1978) 115.
  7. [7]A.N. Skrinskii and Y.M. Shatunov, Precision Measurements of Masses of Elementary Particles Using Storage Rings with Polarized Beams, Sov. Phys. Usp. 32 (1989) 548.DOI
  8. [8]V.A. Sidorov, Results of Experiments at VEPP-2M, in 18th International Conference on High-Energy Physics, (Tbilisi), pp. B13–B18, 1976.
  9. [9]Bayer, V. N. and Khoze, V. A., On determination of transverse polarization of high energy electrons, Yad. Fiz. 9 (1969) 409.
  10. [10]G.Y. Kezerashvili, A.P. Lysenko, Y.M. Shatunov and P.V. Vorobyov, Colliding beam polarization measurement using superconducting helical undulator at the VEPP-2M storage ring, Nucl. Instr. Meth. A 314 (1992) 15.DOI
  11. [11]A.S. Artamonov, S.E. Baru, A.E. Blinov, A.E. Bondar, A.D. Bukin, V.R. Groshev et al., A high precision measurement of the Y-meson mass, Phys. Lett. B 118 (1982) 225.
  12. [12]D. Gustavson, J.J. Murray, T.J. Phillips, R. Schwitters, C.K. Sinclair, J.R. Johnson et al., A Back Scattered Laser Polarimeter for $e^{+} e^{-}$ Storage Rings, Nucl. Instrum. Meth. 165 (1979) 177.
  13. [13]H.D. Bremer, H.C. Dehne, H.C. Lewin, H. Mais, R. Neumann, R. Rossmanith et al., The Petra-Polarimeter, in High-Energy Physics with Polarized Beams and Polarized Targets, C. Joseph and J. Soffer, eds., EXS 38: Experientia Supplementum, (Basel), pp. 469–471, Birkhäuser, 1981, DOI.
  14. [14]D.P. Barber, H.D. Bremer, J. Kewisch, H.C. Lewin, T. Limberg, H. Mais et al., A precision measurement of the Y’ meson mass, Phys. Lett. B 135 (1984) 498.DOI
  15. [15]L. Knudsen, J.P. Koutchouk, M. Placidi, R. Schmidt, M. Crozon, J. Badier et al., First observation of transverse beam polarization in LEP, Physics Letters B 270 (1991) 97.
  16. [16]L. Arnaudon, L. Knudsen, J.P. Koutchouk, R. Olsen, M. Placidi, R. Schmidt et al., Measurement of LEP beam energy by resonant spin depolarization, Physics Letters B 284 (1992) 431.DOI
  17. [17]D.P. Barber, H.D. Bremer, M. Böge, R. Brinkmann, W. Brückner, C. Büscher et al., The HERA polarimeter and the first observation of electron spin polarization at HERA, Nucl. Instr. Meth. A 329 (1993) 79.DOI
  18. [18]V.E. Blinov, T.V. Bedareva, S.A. Zakharov, V.V. Kaminskiy, V.N. Kudryavtsev, N.Y. Muchnoi et al., Status of laser polarimeter at VEPP-4M, J. Inst. 15 (2020) C08024.DOI
  19. [19]M. Woods, M. and SLD collaboration, The scanning Compton polarimeter for the SLD experiment, in arXiv:hep-ex/9611005, (Amsterdam, The Netherlands), Nov., 1996, http://arxiv.org/abs/hep-ex/9611005.
  20. [20]I. Passchier, D.W. Higinbotham, C.W. de Jager, B.E. Norum, N.H. Papadakis and N.P. Vodinas, A Compton backscattering polarimeter for measuring longitudinal electron polarization, Nucl. Instr. Meth. A 414 (1998) 446.
  21. [21]M. Beckmann, A. Borissov, S. Brauksiepe, F. Burkart, H. Fischer, J. Franz et al., The Longitudinal Polarimeter at HERA, Nuclear Instruments and Methods in Physics Research Section A: Accelerators, Spectrometers, Detectors and Associated Equipment 479 (2002) 334.DOI
  22. [22]S. Escoffier, P.Y. Bertin, M. Brossard, E. Burtin, C. Cavata, N. Colombel et al., Accurate measurement of the electron beam polarization in JLab Hall A using Compton polarimetry, Nucl. Instr. Meth. A 551 (2005) 563.DOI
  23. [23]C. Bartels, A. Hartin, C. Helebnrant, D. Käfer and J. List, Precision polarimetry at the ILC: Concepts, simulations and experiments, Nucl. Instr. Meth. A 623 (2010) 570.DOI
  24. [24]V.I. Telhov, Beam energy measurement at linear colliders using spin precession, Phys. Rev. ST Accel. Beams 6 (2003) 091002.
  25. [25]A. Blondel, P. Janot, J. Wenninger, R. Ab̈mann, S. Aumon, P. Azzurri et al., Polarization and Centre-of-mass Energy Calibration at FCC-ee, arXiv: 1909.12245 [hep-ex, physics:physics] (2019) .arxiv.org/abs/1909.12245
  26. [26]S. Nikitin, Opportunities to obtain polarization at CEPC, Int. J. Mod. Phys. A 34 (2019) 1940004.DOI
  27. [27]W. Xia, J. Gao, Y. Wang and D. Wang, CEPC Z-pole polarization design studies, Int. J. Mod. Phys. A 36 (2021) 2142003.DOI
  28. [28]A. Accardi, J.L. Albacete, M. Anselmino, N. Armesto, E.C. Aschenauer, A. Bacchetta et al., Electron-Ion Collider: The next QCD frontier, Eur. Phys. J. A 52 (2016) 268.arxiv.org/abs/1212.1701
  29. [29]J. Skaritka, E. Wang, F. Willeke, R. Lambiase, W. Lui, V. Pitsyn et al., Conceptual design of a Polarized Electron Ion Collider at Brookhaven National Laboratory, in Proceedings of XVII International Workshop on Polarized Sources, Targets & Polariemtry - PoS(PSTP2017), vol. 324, p. 015, SISSA Medialab, June, 2018, DOI.
  30. [30]J. Herrmann and V.C. Zhukovskii, Compton Scattering and Induced Compton Scattering in a Constant Electromagnetic Field, Annalen der Physik 482 (1972) 349.DOI
  31. [31]E.V. Abakumova, M.N. Achasov, D.E. Berkaev, V.V. Kaminsky, N.Y. Muchnoi, E.A. Perevedentsev et al., Backscattering of Laser Radiation on Ultrarelativistic Electrons in a Transverse Magnetic Field: Evidence of MeV-Scale Photon Interference, Phys. Rev. Lett. 110 (2013) 140402.pubmed.ncbi.nlm.nih.gov/25166963
  32. [32]N. Muchnoi, H.J. Schreiber and M. Viti, ILC beam energy measurement by means of laser Compton backscattering, Nucl. Instr. Meth. A 607 (2009) 340.DOI
  33. [33]M.N. Achasov and N.Y. Muchnoi, C.M.S. energy calibration in BES-III and VEPP-2000 experiments, J. Inst. 15 (2020) C07005.DOI
  34. [34]V.B. Berestetskii, E.M. Lifshitz and L.P. Pitaevskii, Quantum Electrodynamics: Volume 4, Butterworth-Heinemann (Jan., 1982).
  35. [35]I.B. Mordechai and G. Alexander, A Transverse Polarimeter for a Linear Collider of 250 GeV e± Beam Energy, in Helmholtz Alliance Linear Collider Forum, pp. 577–590, 2010.
  36. [36]A. Abada, M. Abbrescia, S.S. AbdusSalam, I. Abdyukhanov, J. Abel­leira Fernandez, A. Abramov et al., FCC-ee: The Lepton Collider, Eur. Phys. J. Spec. Top. 228 (2019) 261.
  37. [37]N. Muchnoi, Performance of the FCC-ee Polarimeter, in Joint Accelerator Conferences Website, pp. 157–164, JACOW Publishing, Geneva, Switzerland, Apr., 2019, DOI.DOI
  38. [38]Y. Honda, N. Sasao, S. Araki, Y. Higashi, T. Okugi, T. Taniguchi et al., Upgraded laser wire beam profile monitor, Nucl. Instr. Meth. A 538 (2005) 100.DOI

Paper details

vlnc:W_pyjysyqt
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