betaFeedback

WorksPhysics and Astronomy

Coherent γ-ray Generation By Partially Stripped Ion Beams

Abstract

We describe a scheme for generation of coherent γγ-rays by backscattering intense visible laser light from a beam of partially stripped ions. The scheme is similar in principle to the proposed Gamma Factory at CERN, with the important difference that the scattering becomes coherent as a result of a collective instability which microbunches the ions. This instability is analogous to that which occurs in high-gain free electron lasers (FELs). The scheme potentially offers a route to a source of tuneable, coherent γγ-rays, opening up a wide range of possible new applications and opportunities. We find that the parameter requirements for realization of coherent γγ-ray generation regime are considerably more stringent than those proposed for the Gamma Factory, requiring significant increases in the pump laser intensity and possibly the ion beam current.

Introduction

The generation of bright, tunable γ\gamma-ray sources represents a longstanding goal in accelerator and beam physics due to their potential impact across materials science, nuclear physics and particle physics. Recent advances in relativistic ion beams have motivated renewed interest in schemes based interactions between ion beams and lasers, most notably the Gamma Factory project at CERN, which proposes to produce high-energy photons via Doppler upshifting of laser light scattered by partially stripped ions circulating in storage rings [1, 2, 3, 4]. Although this approach could potentially produce intense γ\gamma-ray beams, the emitted radiation is fundamentally incoherent, limiting its coherence properties and spectral brightness.

A natural extension of these concepts is the pursuit of methods for production of a laser-like beam of coherent γ\gamma-rays, analogous to the operation of free-electron lasers (FELs) operating in the x-ray regime [5, 6]. In FELs, collective instabilities driven by the interaction between a relativistic electron beam and a radiation field lead to microbunching of the electron beam and exponential amplification of coherent radiation [5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15]. An analogous mechanism exists in atomic and ionic systems in the form of the collective atomic recoil laser (CARL), in which the interplay of optical fields and particle motion leads to self-organization via microbunching of the atoms or ions and consequently coherent scattering of the radiation field [16, 17, 18, 19].

In this work, we investigate the possibility of extending CARL-like dynamics to relativistic beams of partially stripped ions, enabling coherent γ\gamma-ray generation via a collective instability. By considering the interaction of a counter-propagating laser field with a relativistic ion beam, we show that the system can be mapped to a CARL configuration in a suitable co-moving frame, where the scattered and pump fields become quasi-degenerate. In this regime, interference between the fields induces a periodic modulation of the ion density, which in turn enhances coherent backscattering, giving rise to an exponential gain of the γ\gamma-ray field.

We derive a self-consistent theoretical model describing this interaction and analyze the conditions required to achieve the collective, high-gain regime. Particular attention is given to the scaling of the key dimensionless parameter governing the instability and to its dependence on experimentally accessible quantities such as laser intensity, ion beam current, and interaction time. Using parameters relevant to current Gamma Factory proposals, we assess the feasibility of reaching the coherent regime and identify the primary limitations imposed by realistic beam and laser configurations.

Our results indicate that, although coherent γ\gamma-ray generation via partially stripped ion beams is, in principle, achievable, the conditions to be met for its realization are significantly more stringent than those needed for incoherent emission. In particular, significant increases in laser intensity and/or beam current are necessary to trigger the collective instability. These findings will guide future experimental efforts aimed at bridging the gap between incoherent γ\gamma-ray sources and fully coherent laser-like γ\gamma-ray beams.

COHERENT γ\gamma-RAYS FROM ION BEAMS VIA A COLLECTIVE INSTABILITY

Collective Atomic Recoil Laser (CARL)

The collective atomic recoil laser (CARL) model [16, 17, 18, 19, 20, 21, 22, 23] describes collective Rayleigh scattering of light from cold atoms or ions. It was originally developed by assuming a cold thermal atomic gas with zero average velocity and quasi-degenerate pump and scattered optical fields, as would be expected for elastic Rayleigh scattering. It describes a process where interference between pump and scattered fields creates a spatially periodic optical potential which causes bunching of the atoms. These bunched atoms in turn scatter the optical pump beam coherently to amplify the backscattered beam via a collective instability in which many atoms interact collectively via the common optical fields with which they interact.

The CARL concept was extended to include a relativistic beam of ions in [24]. The system behaves very similarly to the case of a zero average-velocity thermal gas of ions when we transform to a frame of reference co-moving or almost co-moving with the ions, as shown schematically in Figure 1. However, in the lab frame the frequencies of the pump field and scattered field are now very different due to the large Doppler shift involved.

Schematic 1D representation of the $\gamma$-ray generation process in the laboratory and moving frames

Figure 1. Schematic 1D representation of the γ\gamma-ray generation process in (a) the laboratory frame (S) and (b) in the moving frame (S′).

Here we reconsider the relativistic CARL model, as described in [24], and use it to determine the requirements for performing coherent γ\gamma-ray generation by a beam of partially stripped ions like that proposed for the Gamma Factory at CERN.

Model & Assumptions

We consider a mono-energetic beam of two-level ions having a proper transition frequency ω0′\omega'_0. In the laboratory reference frame, SS, the ions interact with a strong counter-propagating pump field

E2=E02x^exp⁡[i(k2z+ω2t)].E_2=E_{02}\hat{x}\exp[i(k_2z+\omega_2t)].

where k2=ω2/ck_2=\omega_2/c, and by a weak co-propagating probe field

E1=E01x^exp⁡[i(k1z−ω1t)].E_1=E_{01}\hat{x}\exp[i(k_1z-\omega_1t)].

We consider a reference frame S′S' moving at a velocity vr=cβr=c(ω1−ω2)/(ω1+ω2)v_r=c\beta_r=c(\omega_1-\omega_2)/(\omega_1+\omega_2) where the two frequencies coincide, as shown in Figure 1. This can be seen using the Lorentz transformations

ω2,1′=γrω2,1(1±βr).(1)\omega'_{2,1}=\gamma_r\omega_{2,1}(1\pm\beta_r). \tag*{(1)}

If ω1′=ω2′=ω′\omega'_1=\omega'_2=\omega', then ω2(1+βr)=ω1(1−βr)\omega_2(1+\beta_r)=\omega_1(1-\beta_r) and ω1=ω2(1+βr)/(1−βr)=(1+βr)2γr2ω2\omega_1=\omega_2(1+\beta_r)/(1-\beta_r)=(1+\beta_r)^2\gamma_r^2\omega_2. The resonance condition ω′=ω0′\omega'=\omega'_0 implies ω2,1=ω0′/[γr(1±βr)]\omega_{2,1}=\omega'_0/[\gamma_r(1\pm\beta_r)].

CARL in the moving frame (S′S')

We assume the ion motion to be non-relativistic in the moving frame S′S'. The system is then represented by the following Hamiltonian

H=∑j=1N{ℏω0′2σ3j+pj′22m+ℏg′[a1′†σj−e−i(k′zj′−ω′t′)+a2′†σj−ei(k′zj′+ω′t′)+h.c.]}.(2)H=\sum_{j=1}^{N}\left\{\frac{\hbar\omega'_0}{2}\sigma_{3j}+\frac{p_j'^2}{2m}+\hbar g'\left[a_1'^\dagger\sigma_j^-e^{-i(k'z'_j-\omega't')}+a_2'^\dagger\sigma_j^-e^{i(k'z'_j+\omega't')}+\mathrm{h.c.}\right]\right\}. \tag*{(2)}

where wavenumber k′=ω′/ck'=\omega'/c and coupling rate g′=μ[ω′/(2ℏϵ0V′)]1/2g'=\mu[\omega'/(2\hbar\epsilon_0V')]^{1/2}, where V′V' is the quantization volume of the radiation field and μ\mu is the electric dipole matrix element of the two-level transition. The internal dynamics of the two-level ions are described by the operators σ3j=∣ej⟩⟨ej∣−∣gj⟩⟨gj∣\sigma_{3j}=|e_j\rangle\langle e_j|-|g_j\rangle\langle g_j|, σj+=∣ej⟩⟨gj∣\sigma_j^+=|e_j\rangle\langle g_j| and σj−=∣gj⟩⟨ej∣\sigma_j^-=|g_j\rangle\langle e_j|. Furthermore, we also consider the dynamics of external degrees of freedom, where zj′z'_j and pj′p'_j are operators. The Heisenberg equations are:

dzj′dt=pj′m,(3)\frac{dz'_j}{dt}=\frac{p'_j}{m}, \tag*{(3)}
dpj′dt=iℏk′g′[a1′†σj−e−i(k′zj′−ω′t′)−a2′†σj−ei(k′zj′+ω′t′)−h.c.].(4)\frac{dp'_j}{dt}=i\hbar k'g'\left[a_1'^\dagger\sigma_j^-e^{-i(k'z'_j-\omega't')}-a_2'^\dagger\sigma_j^-e^{i(k'z'_j+\omega't')}-\mathrm{h.c.}\right]. \tag*{(4)}
dσj−dt=−iω0′σj−+ig′σ3j{a1′ei(k′z′−ω′t)+a2′e−i(k′z′+ω′t)}(5)\frac{d\sigma_j^-}{dt}=-i\omega'_0\sigma_j^-+ig'\sigma_{3j}\left\{a'_1e^{i(k'z'-\omega't)}+a'_2e^{-i(k'z'+\omega't)}\right\} \tag*{(5)}
dσ3jdt=2ig′[a1′†σj−e−i(k′z′−ω′t)+a2′†σj−ei(k′z′+ω′t)−h.c.](6)\frac{d\sigma_{3j}}{dt}=2ig'\left[a_1'{}^\dagger\sigma_j^-e^{-i(k'z'-\omega't)}+a_2'{}^\dagger\sigma_j^-e^{i(k'z'+\omega't)}-\mathrm{h.c.}\right] \tag*{(6)}
da1′dt=−ig′∑j=1Nσj−e−i(k′z′−ω′t)(7)\frac{da'_1}{dt}=-ig'\sum_{j=1}^{N}\sigma_j^-e^{-i(k'z'-\omega't)} \tag*{(7)}

where we have assumed that the pump field is sufficiently strong to be represented by a constant real number a2′a'_2. Furthermore, we neglect the population of the excited state, assuming σ3j≈−1\sigma_{3j}\approx-1. Introducing the coherence variable σj=σj−ei(k′z′+ω′t)\sigma_j=\sigma_j^-e^{i(k'z'+\omega't)}, the equations become

dzj′dt=pj′m(8)\frac{dz'_j}{dt}=\frac{p'_j}{m} \tag*{(8)}
dpj′dt=iℏk′g′[a1′†σje−2ik′z′−a2′†σj−h.c.](9)\frac{dp'_j}{dt}=i\hbar k'g'\left[a'_1{}^\dagger\sigma_je^{-2ik'z'}-a'_2{}^\dagger\sigma_j-\mathrm{h.c.}\right] \tag*{(9)}
dσjdt=i(Δ′+k′pj′m)σj−ig′{a1′e2ik′z′+a2′}−Γ′2σj(10)\frac{d\sigma_j}{dt}=i\left(\Delta'+\frac{k'p'_j}{m}\right)\sigma_j-ig'\left\{a'_1e^{2ik'z'}+a'_2\right\}-\frac{\Gamma'}{2}\sigma_j \tag*{(10)}
da1′dt=−ig′∑j=1Nσje−2ik′z′(11)\frac{da'_1}{dt}=-ig'\sum_{j=1}^{N}\sigma_je^{-2ik'z'} \tag*{(11)}

where Δ′=ω′−ω0′\Delta'=\omega'-\omega'_0 is the detuning between the radiation fields and the transition, and we have added the spontaneous emission decay term −(Γ′/2)σj-(\Gamma'/2)\sigma_j, with Γ=μ2k′3/2πϵ0ℏ\Gamma=\mu^2k'^3/2\pi\epsilon_0\hbar as the rate of spontaneous decay. In order to eliminate the internal degrees of freedom we neglect the Doppler shift k′pj′/mk'p'_j/m and Γ′\Gamma' with respect to the detuning Δ′\Delta' and neglect a1′a'_1 with respect to the pump field a2′a'_2. For times much larger than 1/Γ1/\Gamma the time derivative can be neglected, σ˙j≈0\dot{\sigma}_j\approx0, and we obtain σj≈g′a2′/Δ′\sigma_j \approx g'a'_2/\Delta' and

dθj′dt′=2k′pj′m(12)\frac{d\theta'_j}{dt'} = 2k'\frac{p'_j}{m} \tag*{(12)}
dpj′dt′=−ℏk′g′2a2′Δ′[a′eiθj′+a′†e−iθj′](13)\frac{dp'_j}{dt'} = -\hbar k'\frac{g'^2a'_2}{\Delta'}\left[a'e^{i\theta'_j}+a'^\dagger e^{-i\theta'_j}\right] \tag*{(13)}
da′dt′=g′2a2′Δ′∑j=1Ne−iθj′.(14)\frac{da'}{dt'} = \frac{g'^2a'_2}{\Delta'}\sum_{j=1}^{N}e^{-i\theta'_j}. \tag*{(14)}

where we have defined θ′=2k′z′\theta'=2k'z' and a′=ia1′a'=ia'_1. These are the CARL equations in the moving frame S′S'.

Transformation to the laboratory frame (S)

We now transform Eqs. (12)–(14) into the laboratory frame SS. The phase θj′\theta'_j transforms as

θj′=2k′z′=k1(zj−ct)+k2(zj+ct)=(k1+k2)(zj−cβrt)=θj(15)\theta'_j=2k'z'=k_1(z_j-ct)+k_2(z_j+ct)=(k_1+k_2)(z_j-c\beta_rt)=\theta_j \tag*{(15)}

The momentum pj′p'_j transforms as

pj′=γr(pj−βrEj/c)=mγrγj(vj−vr)(16)p'_j=\gamma_r(p_j-\beta_rE_j/c)=m\gamma_r\gamma_j(v_j-v_r) \tag*{(16)}

since pj=mγjvjp_j=m\gamma_jv_j and Ej=mc2γjE_j=mc^2\gamma_j, with γj=(1−βj2)−1/2\gamma_j=(1-\beta_j^2)^{-1/2} and βj=vj/c\beta_j=v_j/c. Then, since dt′=dt/γrdt'=dt/\gamma_r,

dθjdt=2k2(1+βr)γrγj(vj−vr)(17)\frac{d\theta_j}{dt}=2k_2(1+\beta_r)\gamma_r\gamma_j(v_j-v_r) \tag*{(17)}

We define ηj=γrγj(βj−βr)\eta_j=\gamma_r\gamma_j(\beta_j-\beta_r), so that

dθjdt=2ω2(1+βr)ηj(18)\frac{d\theta_j}{dt}=2\omega_2(1+\beta_r)\eta_j \tag*{(18)}

We observe that in the non relativistic CARL (βj≪1\beta_j\ll1 and βr≪1\beta_r\ll1), ηj=(vj−vr)/c\eta_j=(v_j-v_r)/c, whereas in the ultra-relativistic FEL limit, (βj,r≈1−1/2γj,r2(\beta_{j,r}\approx1-1/2\gamma_{j,r}^2 and γj∼γr)\gamma_j\sim\gamma_r), ηj=2(γj−γr)/γr\eta_j=2(\gamma_j-\gamma_r)/\gamma_r.

Let now consider the coupling constant g′=μ[ω′/(2ℏϵ0V′)]1/2g'=\mu[\omega'/(2\hbar\epsilon_0V')]^{1/2}. Since V′=γrVV'=\gamma_rV and ω′=γrω2,1(1+βr)\omega'=\gamma_r\omega_{2,1}(1+\beta_r), g′=g21+βr=g11−βrg'=g_2\sqrt{1+\beta_r}=g_1\sqrt{1-\beta_r}, where g1,2=μ[ω1,2/(2ℏϵ0V)]1/2g_{1,2}=\mu[\omega_{1,2}/(2\hbar\epsilon_0V)]^{1/2}.

The detuning transforms as Δ′=γr(1+βr)ω2−ω0′\Delta'=\gamma_r(1+\beta_r)\omega_2-\omega'_0. We do not write it in the laboratory values, since it refers to the detuning with respect to the proper transition frequency ω0′\omega'_0.

Let now consider the transformations for a′a' and a2′a'_2. Since a′†a′=ϵ0E0′2V′/2ℏωa'^\dagger a'=\epsilon_0E_0'^2V'/2\hbar\omega, and since E1′=γr(1−βr)E1E'_1=\gamma_r(1-\beta_r)E_1 and E2′=γr(1+βr)E2E'_2=\gamma_r(1+\beta_r)E_2, we have

a′†a′=a†aγr2(1−βr)(19)a'^\dagger a'=a^\dagger a\gamma_r^2(1-\beta_r) \tag*{(19)}
a2′†a2′=a2†a2γr2(1+βr)(20)a_2'^\dagger a'_2=a_2^\dagger a_2\gamma_r^2(1+\beta_r) \tag*{(20)}
Equation not extracted
(21)

so that

a′=a1+βr(22)a'=\frac{a}{\sqrt{1+\beta_r}} \tag*{(22)}
a2′=a21−βr=a2γr1+βr(23)a'_2=\frac{a_2}{\sqrt{1-\beta_r}}=a_2\gamma_r\sqrt{1+\beta_r} \tag*{(23)}

Then, Eqs. (12)–(14) transform into:

dθjdt=2ω2(1+βr)ηj(24)\frac{d\theta_j}{dt}=2\omega_2(1+\beta_r)\eta_j \tag*{(24)}
dηjdt=−ℏω2mc2g22a2(1+βr)2γrΔ′[aeiθj+a†e−iθj](25)\frac{d\eta_j}{dt}=-\frac{\hbar\omega_2}{mc^2}\frac{g_2^2a_2(1+\beta_r)^2\gamma_r}{\Delta'}\left[ae^{i\theta_j}+a^\dagger e^{-i\theta_j}\right] \tag*{(25)}
dadt=g22a2(1+βr)2NΔ′⟨exp⁡(−iθ)⟩,(26)\frac{da}{dt}=\frac{g_2^2a_2(1+\beta_r)^2N}{\Delta'}\langle\exp(-i\theta)\rangle, \tag*{(26)}

where ⟨exp⁡(−iθ)⟩=(1/N)∑j=1Nexp⁡(−iθj)\langle\exp(-i\theta)\rangle=(1/N)\sum_{j=1}^{N}\exp(-i\theta_j).

CARL scaling

We introduce dimensionless time τ=2ω2(1+βr)ρt\tau= 2\omega_2(1+\beta_r)\rho t and dimensionless momentum pˉj=ηj/ρ\bar{p}_j = \eta_j/\rho,

dθjdτ=pˉj(27)\frac{d\theta_j}{d\tau} = \bar{p}_j \tag*{(27)}
dpˉjdτ=−C1ρ2[aeiθj+a†e−iθj],(28)\frac{d\bar{p}_j}{d\tau} = -\frac{C_1}{\rho^2}\left[a e^{i\theta_j}+a^\dagger e^{-i\theta_j}\right], \tag*{(28)}
dadt=C2ρ⟨exp⁡(−iθ)⟩,(29)\frac{da}{dt} = \frac{C_2}{\rho}\langle\exp(-i\theta)\rangle, \tag*{(29)}

where C1=[ℏg22a2(1+βr)γr]/(2mc2Δ′)C_1 = [\hbar g_2^2a_2(1+\beta_r)\gamma_r]/(2mc^2\Delta') and C2=[g22a2(1+βr)N]/(2ω2Δ′)C_2 = [g_2^2a_2(1+\beta_r)N]/(2\omega_2\Delta'). By defining A=ρa/C2=C1a/ρ2A = \rho a/C_2 = C_1a/\rho^2 we obtain

ρ=(C1C2)1/3=[ℏg24a22(1+βr)2γrN4mc2ω2Δ′2]1/3,(30)\rho= (C_1C_2)^{1/3} = \left[\frac{\hbar g_2^4a_2^2(1+\beta_r)^2\gamma_rN}{4mc^2\omega_2\Delta'^2}\right]^{1/3}, \tag*{(30)}
A=C1C2aρ=ℏω2γrmc2Nρa(31)A = \sqrt{\frac{C_1}{C_2}}\frac{a}{\sqrt{\rho}} = \sqrt{\frac{\hbar\omega_2\gamma_r}{mc^2N\rho}}a \tag*{(31)}

and the equations in the standard universal form:

dθjdτ=pˉj(32)\frac{d\theta_j}{d\tau} = \bar{p}_j \tag*{(32)}
dpˉjdτ=−[Aeiθj+A†e−iθj],(33)\frac{d\bar{p}_j}{d\tau} = -\left[Ae^{i\theta_j}+A^\dagger e^{-i\theta_j}\right], \tag*{(33)}
dAdτ=⟨exp⁡(−iθ)⟩.(34)\frac{dA}{d\tau} = \langle\exp(-i\theta)\rangle. \tag*{(34)}

Figure 2 and Figure 3 show the evolution of the ion positions, θj\theta_j, the ion energies, pˉj\bar{p}_j, and the γ\gamma-ray intensity, ∣A∣2|A|^2, as calculated from eq. (32)-(34). Fig. 2 shows that the γ\gamma-ray intensity is exponentially amplified from it’s initially small value before saturating and performing quasi-periodic oscillations. The reason for the exponential amplification can be deduced from Fig. 3, which shows that the amplification of the γ\gamma-ray intensity occurs simultaneously with the development of a strong density modulation in the ion density. It is the development of this density grating that causes the scattering process to change from a single-particle process involving spontaneous scattering to a collective process involving many ions, interacting cooperatively via the common pump and scattered (γ\gamma-ray) field with which they interact. As the ion density grating develops, it scatters the pump field coherently, amplifying the γ\gamma-ray beam and consequently increasing the size of the ion density modulation, resulting in an instability in which the ion density modulation depth and the γ\gamma-ray beam intensity undergo exponential growth simultaneously. This instability is very similar in character to that which underpins high-gain free electron lasers (FELs) [9, 10, 11, 12, 13, 14, 15] and indeed Eq. (32)-(34) are of the same form as the equations describing the high-gain FEL [5].

Dimensionless $\gamma$-ray intensity, $|A|^2$, against interaction time $\tau$

Figure 2. Dimensionless γ\gamma-ray intensity, ∣A∣2|A|^2, against interaction time τ\tau as calculated from Eq. (32)–(34). The points labeled (a), (b) and (c) relate to the phase-space snapshots shown in Figure 3.

Snapshots of ion phase space showing development of ion bunching

Figure 3. Snapshots of ion phase space showing development of ion bunching. The labels (a), (b) and (c) relate to times at which each snapshot occurs (see Figure 2).

In Figs 2 and 3 it has been assumed that the ion beam is initially completely monoenergetic. The existence of the collective instability relies on the ion beam being sufficiently monoenergetic. The threshold condition for the fractional ion beam energy spread, σE/E\sigma_E/E is [8]

σEE<ρ.(35)\frac{\sigma_E}{E} < \rho. \tag*{(35)}

The parameter ρ\rho and the radiated field AA can be written in a different way. We introduce the quantum parameter q=mc2γr/ℏω1q = mc^2\gamma_r/\hbar\omega_1, where ω1=ω2(1+βr)/(1−βr)=ω2γr2(1+βr)2\omega_1 = \omega_2(1+\beta_r)/(1-\beta_r) = \omega_2\gamma_r^2(1+\beta_r)^2. Then

A=aℏω1mc2γr(1+βr)2Nρ=11+βraρˉN(36)A = a\sqrt{\frac{\hbar\omega_1}{mc^2\gamma_r(1+\beta_r)^2N\rho}} = \frac{1}{1+\beta_r}\frac{a}{\sqrt{\bar{\rho}N}} \tag*{(36)}

where ρˉ=qρ\bar{\rho} = q\rho. So, ρˉ(1+βr)2∣A∣2=∣a∣2/N\bar{\rho}(1+\beta_r)^2|A|^2 = |a|^2/N is the average number of photons emitted per ion and ρˉ\bar{\rho} is the quantum efficiency.

Also, 2g2a2=Ω22g_2a_2=\Omega_2, where Ω2=μE02/ℏ\Omega_2=\mu E_0^2/\hbar is the pump Rabi frequency. Introducing the plasma frequency ωp=μ2ω22n/2ϵ0ℏ\omega_p=\sqrt{\mu^2\omega_2^2n/2\epsilon_0\hbar}, where n=N/Vn=N/V is the particle density, we can write

ρ=[Ω2Δ′ωp4ω2(1+βr)]2/3(ℏω2γrmc2)1/3=1q1/3[Ω2Δ′ωp4ω2]2/3.(37)\rho=\left[\frac{\Omega_2}{\Delta'}\frac{\omega_p}{4\omega_2}(1+\beta_r)\right]^{2/3}\left(\frac{\hbar\omega_2\gamma_r}{mc^2}\right)^{1/3}=\frac{1}{q^{1/3}}\left[\frac{\Omega_2}{\Delta'}\frac{\omega_p}{4\omega_2}\right]^{2/3}. \tag*{(37)}

It shows that the parameter ρ\rho scales as γr−1/3\gamma_r^{-1/3}, as predicted by Bonifacio and Barletta [24]. We also have

ρˉ=qρ=[Ω2Δ′ωp4ω2q]2/3=[Ω2Δ′ωpγr4ωrec]2/3(38)\bar{\rho}=q\rho=\left[\frac{\Omega_2}{\Delta'}\frac{\omega_p}{4\omega_2q}\right]^{2/3}=\left[\frac{\Omega_2}{\Delta'}\frac{\omega_p\gamma_r}{4\omega_{\mathrm{rec}}}\right]^{2/3} \tag*{(38)}

where ωrec=ℏω′2/2mc2=ℏω1ω2/2mc2\omega_{\mathrm{rec}}=\hbar\omega'^2/2mc^2=\hbar\omega_1\omega_2/2mc^2 is the recoil frequency. Note that ρˉ\bar{\rho} reduces to the CARL parameter [16] in the non relativistic limit, with ω1=ω2\omega_1=\omega_2 and γr=1\gamma_r=1.

We define detuning as the initial value of pˉj(0)\bar{p}_j(0). For a mono-energetic beam,

δ=pˉ(0)=γrγ0ρ(β0−βr)=γrγ0(1−β0)ρ(ω1+ω2)[ω21+β01−β0−ω1]=γrγ0(1−β0)(1+βr)2ρω1[ω21+β01−β0−ω1](39)\delta=\bar{p}(0)=\frac{\gamma_r\gamma_0}{\rho}(\beta_0-\beta_r)=\frac{\gamma_r\gamma_0(1-\beta_0)}{\rho(\omega_1+\omega_2)}\left[\omega_2\frac{1+\beta_0}{1-\beta_0}-\omega_1\right]=\frac{\gamma_r\gamma_0(1-\beta_0)(1+\beta_r)}{2\rho\omega_1}\left[\omega_2\frac{1+\beta_0}{1-\beta_0}-\omega_1\right] \tag*{(39)}

where we used the expression βr=(ω1−ω2)/(ω1+ω2)\beta_r=(\omega_1-\omega_2)/(\omega_1+\omega_2). If βr≈β0\beta_r\approx\beta_0, then

δ≈12ρω1[ω21+β01−β0−ω1](40)\delta\approx\frac{1}{2\rho\omega_1}\left[\omega_2\frac{1+\beta_0}{1-\beta_0}-\omega_1\right] \tag*{(40)}

The resonance condition δ=0\delta=0 implies ω1=ω2(1+β0)/(1−β0)\omega_1=\omega_2(1+\beta_0)/(1-\beta_0). In this case, the intensity increases exponentially as exp⁡(3τ)\exp(\sqrt{3}\tau). Using t=z/cβ0t=z/c\beta_0, the growth is as exp⁡(z/lg)\exp(z/l_g) where the gain length is

lg=cβ023ω2(1+βr)ρ=cβ0γr23ω′ρ=cβ0γr23ω0′ρ(41)l_g=\frac{c\beta_0}{2\sqrt{3}\omega_2(1+\beta_r)\rho}=\frac{c\beta_0\gamma_r}{2\sqrt{3}\omega'\rho}=\frac{c\beta_0\gamma_r}{2\sqrt{3}\omega'_0\rho} \tag*{(41)}

where we assumed ω′=ω0′\omega'=\omega'_0.

The emitted power is

Pout=ρPbeam∣A∣2(1+βr)2βr=ρˉI(A)Ephoton(eV)∣A∣2(1+βr)2βr(42)P_{out}=\rho P_{beam}|A|^2\frac{(1+\beta_r)^2}{\beta_r}=\bar{\rho}I(A)E_{photon}(eV)|A|^2\frac{(1+\beta_r)^2}{\beta_r} \tag*{(42)}

where Pbeam(W)=I(A)Ebeam(eV)P_{beam}(\mathrm{W})=I(\mathrm{A})E_{beam}(eV) is the beam power and Ephoton(eV)=ℏω1E_{photon}(eV)=\hbar\omega_1 is the photon energy. Notice that ρ(1+βr)2\rho(1+\beta_r)^2 is the maximum energy efficiency, whereas ρˉ(1+βr)2\bar{\rho}(1+\beta_r)^2 is the maximum number of photons emitted per ion. The CARL model is valid in S′S' for (a) Δ′≫Γ′\Delta'\gg\Gamma' and (b) a2′≫a1′a_2' \gg a_1'. This condition in the laboratory frame brings to

P2≫(1−βr1+βr)2Pout.(43)P_2\gg\left(\frac{1-\beta_r}{1+\beta_r}\right)^2P_{out}. \tag*{(43)}

About the particle beam, we take the radius of the beam as rb=(β∗ϵn/γr)1/2r_b=(\beta^*\epsilon_n/\gamma_r)^{1/2}, where β∗\beta^* is the beta function and ϵn\epsilon_n is the normalized emittance. We assume β∗=lg\beta^*=l_g and the beam current ibeam=ecβrnπrb2i_{beam}=ec\beta_r n\pi r_b^2. Hence,

n=ibeamγrecβrπϵnβ∗=ibeamγrecβrπϵn23ω′ρcβrγr=23ibeamω′ec2βr2πϵnρn=\frac{i_{beam}\gamma_r}{ec\beta_r\pi\epsilon_n\beta^*}=\frac{i_{beam}\gamma_r}{ec\beta_r\pi\epsilon_n}\frac{2\sqrt{3}\omega'\rho}{c\beta_r\gamma_r}=\frac{2\sqrt{3}i_{beam}\omega'}{ec^2\beta_r^2\pi\epsilon_n}\rho

By inserting it in the expression of ρ3\rho^3.

ρ3=1q[Ω2Δ′ωp4ω2]2=1q[Ω24Δ′ω2]2μ2ω22ℏϵ0n=1q[Ω2Δ′]2μ2ℏϵ03ibeamω′16ω2ec2βr2πϵnρ\rho^3=\frac{1}{q}\left[\frac{\Omega_2}{\Delta'}\frac{\omega_p}{4\omega_2}\right]^2=\frac{1}{q}\left[\frac{\Omega_2}{4\Delta'\omega_2}\right]^2\frac{\mu^2\omega_2}{2\hbar\epsilon_0}n=\frac{1}{q}\left[\frac{\Omega_2}{\Delta'}\right]^2\frac{\mu^2}{\hbar\epsilon_0}\frac{\sqrt{3}i_{beam}\omega'}{16\omega_2ec^2\beta_r^2\pi\epsilon_n}\rho

and

ρ=μ4cβr[Ω2Δ′]3ibeamω′qω2eπℏϵ0ϵn(44)\rho=\frac{\mu}{4c\beta_r}\left[\frac{\Omega_2}{\Delta'}\right]\sqrt{\frac{\sqrt{3}i_{beam}\omega'}{q\omega_2e\pi\hbar\epsilon_0\epsilon_n}} \tag*{(44)}

REQUIREMENTS FOR COHERENT γ\gamma-RAY GENERATION

The proof of principle experiment at CERN will use a beam of ions 208Pb79+{}^{208}\mathrm{Pb}^{79+}, with a transition 2s→2p1/22s \to2p_{1/2} at 230.81 eV and a lifetime of the excited state τ=76.6\tau= 76.6 ps. The relativistic Lorentz factor is γ=96.3\gamma= 96.3 and the ion mass m=193.687 GeV/c2m = 193.687\,\mathrm{GeV}/c^2. The number of ions per bunch is N=108N = 10^8, the length of the bunch τbeam=213\tau_{\mathrm{beam}} = 213 ps, the normalized emittance ϵn=1.5\epsilon_n = 1.5 mm mrad and the energy spread σE/E=2×10−4\sigma_E/E = 2 \times10^{-4}.

The resonance frequency is ω0′=3.5066×1017 s−1\omega'_0 = 3.5066 \times10^{17}\,\mathrm{s}^{-1}, corresponding to a wavelength λ0′=5.375\lambda'_0 = 5.375 nm. The pump frequency is ω2=ω0′/2γ=1.82×1015 s−1\omega_2 = \omega'_0/2\gamma= 1.82 \times10^{15}\,\mathrm{s}^{-1}, with wavelength λ2=2γλ0′=1035.22\lambda_2 = 2\gamma\lambda'_0 = 1035.22 nm. The output frequency is ω1=2γω0′=6.753×1019 s−1\omega_1 = 2\gamma\omega'_0 = 6.753 \times10^{19}\,\mathrm{s}^{-1} (i.e. ℏω1=44.454\hbar\omega_1 = 44.454 keV). The decay rate is Γ′=1/τ=1.305×1010 s−1\Gamma' = 1/\tau= 1.305 \times10^{10}\,\mathrm{s}^{-1}. A detuning of δa=Δ′/Γ′=103\delta_a = \Delta'/\Gamma' = 10^3 corresponds to a relative frequency shift of about 10−410^{-4}, so it can be neglected in the remaining expressions. Assuming Γ′=μ2ω0′3/3πℏϵ0c3\Gamma' = \mu^2\omega_0'^3/3\pi\hbar\epsilon_0c^3, we evaluate the electric dipole as μ=2.68×10−31\mu= 2.68 \times10^{-31} C m. The other derived parameters are the following:

  • q=mc2γ/ℏω1=4.2×108q = mc^2\gamma/\hbar\omega_1 = 4.2 \times10^8;

  • ωrec=ℏω1ω2/2mc2=(ℏω0′)2/2mc2ℏ=2.08×108 s−1\omega_{\mathrm{rec}} = \hbar\omega_1\omega_2/2mc^2 = (\hbar\omega'_0)^2/2mc^2\hbar= 2.08 \times10^8\,\mathrm{s}^{-1};

  • ωp=μ2ω2n/2ϵ0ℏ=(λ0′/4π)3πcn/γτ=0.265n[m−3] s−1\omega_p = \sqrt{\mu^2\omega_2 n/2\epsilon_0\hbar} = (\lambda'_0/4\pi)\sqrt{3\pi cn/\gamma\tau} = 0.265\sqrt{n[\mathrm{m}^{-3}]}\,\mathrm{s}^{-1};

  • ρ=q−1/3(Ω2/Δ′)2/3(ωp/4ω2)2/3=0.8×10−13(Ω2/Δ′)2/3(n[m−3])1/3\rho= q^{-1/3}(\Omega_2/\Delta')^{2/3}(\omega_p/4\omega_2)^{2/3} = 0.8 \times10^{-13}(\Omega_2/\Delta')^{2/3}(n[\mathrm{m}^{-3}])^{1/3};

  • ρˉ=qρ=3.32×10−5(Ω2/Δ′)2/3(n[m−3])1/3\bar{\rho} = q\rho= 3.32 \times10^{-5}(\Omega_2/\Delta')^{2/3}(n[\mathrm{m}^{-3}])^{1/3};

  • Ω2=μE2/ℏ=2.54×103×E2[V/m] s−1\Omega_2 = \mu E_2/\hbar= 2.54 \times10^3 \times E_2[\mathrm{V/m}]\,\mathrm{s}^{-1};

  • Ω2/Δ′=1.94×10−7×(E2[V/m]/δa)\Omega_2/\Delta' = 1.94 \times10^{-7} \times(E_2[\mathrm{V/m}]/\delta_a);

  • ρ=0.27×10−17×(E2[V/m]/δa)2/3×(n[m−3])1/3\rho= 0.27 \times10^{-17} \times(E_2[\mathrm{V/m}]/\delta_a)^{2/3} \times(n[\mathrm{m}^{-3}])^{1/3};

  • lg=cγ/23ω0′ρ=2.4×10−8/ρl_g = c\gamma/2\sqrt{3}\omega'_0\rho= 2.4 \times10^{-8}/\rho m;

  • tg=γ/23ω0′ρ=8.0×10−17/ρt_g = \gamma/2\sqrt{3}\omega'_0\rho= 8.0 \times10^{-17}/\rho s.

The laser delivers a pulse of energy 5mJ for τL=2.8\tau_L = 2.8ps, i.e. a power PL=1.8P_L = 1.8GW. Since E2=2IL/ϵ0c=27×IL1/2[W/m2]E_2 = \sqrt{2I_L/\epsilon_0c} = 27 \times I_L^{1/2}[\mathrm{W/m^2}] and the laser beam has a transverse size σT=0.65\sigma_T = 0.65 mm, IL=PL/πσT2∼1015W/m2I_L = P_L/\pi\sigma_T^2 \sim10^{15}\mathrm{W/m^2} and E2∼109V/mE_2 \sim10^9\mathrm{V/m}. If n∼1018m−3n \sim10^{18}\mathrm{m}^{-3}, we have ρ≈0.3×10−5×δa−2/3\rho\approx0.3 \times10^{-5} \times\delta_a^{-2/3}. For δa∼103\delta_a \sim10^3 we obtain ρ∼10−8\rho\sim10^{-8}, ρˉ∼4\bar{\rho} \sim4 and a gain length lg∼2.4l_g \sim2.4m.

We can obtain a better estimate of ρ\rho using the self-consistent expression (44). Assuming a beam current ibeam=eN/τbeam=0.075i_{\mathrm{beam}} = eN/\tau_{\mathrm{beam}} = 0.075 A, we obtain, for the parameters above, ρ=0.4×10−6/δa\rho= 0.4 \times10^{-6}/\delta_a.

Based on the parameters considered above, the scattering process will be spontaneous in nature and will not reach the collective regime of exponential growth of the γ\gamma-ray beam intensity and ion bunching due to the condition for a monoenergetic ion beam, Eq. (35), not being satisfied, as typical values of ρ≈10−8\rho\approx10^{-8} and σE/E≈10−4\sigma_E/E \approx10^{-4}. From the definition of ρ\rho in Eq. (44), it can be seen that

ρ∝(ILibeam)1/2(45)\rho\propto(I_L i_{\mathrm{beam}})^{1/2} \tag*{(45)}

so in order to satisfy Eq. (35) for the ion-beam proposed for the Gamma Factory, the laser intensity ILI_L would require to be increased by a factor of ∼108\sim10^8 to IL∼1023Wm−2=1019Wcm−2I_L \sim10^{23}\mathrm{Wm}^{-2} = 10^{19}\mathrm{Wcm}^{-2}. The requirements on the pump intensity could be relaxed somewhat in principle if the ion beam current was increased by increasing the number of ions in each bunch or reducing the bunch duration. In addition, the pump laser duration should be sufficient to reach the collective regime of scattering with exponential growth of the γ\gamma-ray beam intensity and strong ion microbunching (approximately points (b) and (c) in Figs 2 and 3), which requires τ>1\tau> 1 i.e. tint>tgt_{\mathrm{int}} > t_g. For ρ≈10−4\rho\approx10^{-4}, this condition corresponds to tint>1t_{\mathrm{int}} > 1ps.

CONCLUSIONS

We have presented a theoretical study of a scheme for generation of coherent γ\gamma-rays by scattering an intense laser from a beam of partially stripped ions traveling at relativistic velocities. The scheme is based on the proposed Gamma Factory at CERN, which aims to generate incoherent γ\gamma-rays using a similar approach. The essential difference between the Gamma Factory and the scheme described here for coherent γ\gamma-ray generation is that coherent γ\gamma-ray generation relies on a collective instability involving back-action of the spontaneously generated γ\gamma-rays to produce bunching of the ions, resulting in exponential amplification of the γ\gamma-rays and the ion density modulation. The requirements for realization of this collective instability and the coherent γ\gamma-ray generation regime are, unsurprisingly, more stringent than those required for the Gamma Factory proposal for incoherent γ\gamma-rays generated spontaneously. In order to attain a regime of coherent γ\gamma-ray generation using the partially stripped ion beams proposed at CERN will require significant increases in the pump laser intensity and possibly the ion beam current, relative to the values currently proposed for the Gamma Factory.

References

  1. [1]M. W. Krasny, The Gamma Factory proposal for CERN, arXiv:1511.07794 [hep-ex] (2015).arxiv.org/abs/1511.07794
  2. [2]D. Budker, J. R. Crespo López-Urrutia, A. Derevianko, V. V. Flambaum, M. W. Krasny, A. Petrenko, S. Pustelny, A. Surzhykov, V. A. Yerokhin, and M. Zolotorev, Atomic Physics Studies at the Gamma Factory at CERN, Annalen Phys. 532, 2000204 (2020), arXiv:2003.03855 [physics.atom-ph].DOI
  3. [3]D. Budker et al., Expanding Nuclear Physics Horizons with the Gamma Factory, arXiv:2106.06584 [nucl-ex] (2021).DOI
  4. [4]D. Budker, M. W. Krasny, and Y. Dutheil, Gamma factory: The light into the future, International Journal of Modern Physics A 0, 2649003 (0), https://doi.org/10.1142/S0217751X26490035.
  5. [5]B. W. J. McNeil and N. R. Thompson, X-ray free-electron lasers, Nature Photonics 4, 814 (2010).
  6. [6]C. Pellegrini, A. Marinelli, and S. Reiche, The physics of x-ray free-electron lasers, Rev. Mod. Phys. 88, 015006 (2016).DOI
  7. [7]A. Kondratenko and E. Saldin, Generating of coherent radiation by a relativistic electron beam in an undulator, Part. Accel. 10, 207 (1980).
  8. [8]R. Bonifacio, C. Pellegrini, and L. Narducci, Collective instabilities and high-gain regime in a free electron laser, Optics Communications 50, 373 (1984).
  9. [9]P. Emma et al., First lasing and operation of an ångstrom-wavelength free-electron laser, Nature Photonics 4, 641 (2010).
  10. [10]M. Altarelli, R. Brinkmann, and M. Chergui, The european x-ray free-electron laser, Technical design report (2007).
  11. [11]B. Faatz, N. Baboi, V. Ayyazyan, V. Balandin, W. Decking, S. Duesterer, H. Eckoldt, J. Feldhaus, N. Golubeva, M. Koerfer, et al., Flash ii: a seeded future at flash, Proceedings of IPAC, Kyoto, Japan (2010).DOI
  12. [12]T. Shintake, H. Tanaka, T. Hara, T. Tanaka, K. Togawa, M. Yabashi, Y. Otake, Y. Asano, T. Bizen, T. Fukui, et al., A compact free-electron laser for generating coherent radiation in the extreme ultraviolet region, Nature Photonics 2, 555 (2008).
  13. [13]T. Shintake et al., Status report on japanese xfel construction project at spring-8, in Proceedings of the 2010 International Particle Accelerator Conference (IEEE, 2010).
  14. [14]B. D. Patterson et al., Coherent science at the swissfel x-ray laser, New Journal of Physics 12, 035012 (2010).DOI
  15. [15]C. Kim et al., Review of technical achievements in pal-xfel, AAPS Bulletin 10.1007/s43673-022-00045-4 (2022).
  16. [16]R. Bonifacio and L. De Salvo, Collective atomic recoil laser (carl) optical gain without inversion by collective atomic recoil and self-bunching of two-level atoms, Nuclear Instruments and Methods in Physics Research Section A: Accelerators, Spectrometers, Detectors and Associated Equipment 341, 360 (1994).
  17. [17]R. Bonifacio, L. De Salvo, L. M. Narducci, and E. J. D’Angelo, Exponential gain and self-bunching in a collective atomic recoil laser, Phys. Rev. A 50, 1716 (1994).
  18. [18]R. Bonifacio, G. R. M. Robb, and B. W. J. McNeil, Propagation, cavity, and doppler-broadening effects in the collective atomic recoil laser, Phys. Rev. A 56, 912 (1997).
  19. [19]N. Piovella, R. Bonifacio, B. McNeil, and G. Robb, Superradiant light scattering and grating formation in cold atomic vapours, Optics Communications 187, 165 (2001).
  20. [20]C. von Cube, S. Slama, D. Krusz, C. Zimmermann, P. W. Courteille, G. R. M. Robb, N. Piovella, and R. Bonifacio, Self-synchronization and dissipation-induced threshold in collective atomic recoil lasing, Phys. Rev. Lett. 93, 083601 (2004).
  21. [21]A. T. Gisbert and N. Piovella, Multimode collective atomic recoil lasing in free space, Atoms 8, 93 (2020).
  22. [22]S. Slama, S. Bux, G. Krenz, C. Zimmermann, and P. W. Courteille, Superradiant rayleigh scattering and collective atomic recoil lasing in a ring cavity, Phys. Rev. Lett. 98, 053603 (2007).
  23. [23]H. Tomczyk, D. Schmidt, C. Georges, S. Slama, and C. Zimmermann, Stability diagram of the collective atomic recoil laser with thermal atoms, Phys. Rev. A 91, 063837 (2015).
  24. [24]R. Bonifacio, L. De Salvo, and W. Barletta, Relativistic theory of the collective atomic recoil laser, Nuclear Instruments and Methods in Physics Research Section A: Accelerators, Spectrometers, Detectors and Associated Equipment 384, 337 (1997).

Paper details

vlnc:W_9642y2k4
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