logo

Scheme for generating 1 nm X-ray beams carrying orbital angular momentum at the SXFEL

SYNCHROTRON RADIATION TECHNOLOGY AND APPLICATIONS

Scheme for generating 1 nm X-ray beams carrying orbital angular momentum at the SXFEL

He-Ping Geng
Jian-Hui Chen
Zhen-Tang Zhao
Nuclear Science and TechniquesVol.31, No.9Article number 88Published in print 01 Sep 2020Available online 28 Aug 2020
59800

Optical vortices have the main features of helical wavefronts and spiral phase structures, and carry orbital angular momentum (OAM). This special structure of visible light has been produced and studied for various applications. These notable characteristics of photons were also tested in the extreme-ultraviolet and X-ray regimes. In this article, we simulate the use of a simple afterburner configuration by directly adding helical undulators after the SASE undulators with the Shanghai Soft X-ray FEL to generate high intensity X-ray vortices with wavelengths 1 nm. Compared to other methods, this approach is easier to implement, cost-effective, and more efficient.

X-rayOrbital angular momentum (OAM)Synchrotron light sourceFree-electron laser (FEL)

1 Introduction

Structured light, whose phase structures rotate when they propagate through space, has been a topic of intense research. With the development of tools and technologies for creating and detecting structured light, applications have begun to appear steadily [1]. Among these studies, the optics and laser communities are interested in phase singularity[2] and visible photons carrying orbital angular momentum (OAM)[3], to provide new explanations of optical effects and develop distinct applications ranging from manipulations of small particles, driving micro-machines, and quantum computation, to optical communications[4-14].

The noteworthy attributes of photons were also demonstrated in the extreme-ultraviolet and X-ray or ever γ ray[15] regions, pioneered by Peele and co-workers[16, 17]. Triggered by this pioneering effort, various generation methods based on particle accelerators[18-25], high-power lasers[26-29], and plasma[30, 31], as well as possible novel applications[32-38], are growing steadily.

Undulators are the workhorses in third-generation synchrotron radiation light sources and free-electron lasers (FELs) for generating X-ray beams. Sasaki and McNulty[39] were the first to indicate that when electrons traveled along a spiral path in helical undulators, the harmonics produced by them carried a spiral phase. Therefore, it is proposed that the helical undulator could be used as a source for generating intense X-rays carrying OAM light (XOAM). Bahrdt and coworkers at BESSY-II[40] first measured the special helical structure in spontaneous undulator emission. This was later measured in the XUV (X-ray and extreme ultraviolet) region at UVSOR-III[41] by Tatsuo and his coworkers, and in coherent radiation emitted from energy and density modulated (bunched) electron beams[42] by the US group at SLAC. In 2014, Hemsing and coworkers described a simpler afterburner configuration based on the HGHG setup to generate OAM light[43]. Based on Hemsing’s HGHG scheme, Primoz Rebernik Ribic et al. in 2014 proposed the use of an optical phase mask to produce a transverse phase modulation in the incident seed laser, which could also produce an XUV or X-ray optical vortex[44]. Three years later, in 2017, they demonstrated two techniques to obtain intense femtosecond coherent optical vortices at the FERMI FEL in the extreme ultraviolet[45]. The first method used the helical undulator to produce spiral beams at the second harmonic, while the second method relied on a spiral zone plate by putting it directly into the propagation path of the transverse Gaussian optical beam.

In this study, we simulate the use of a simple afterburner configuration by directly adding helical undulators after the SASE undulators at the Shanghai Soft X-ray FEL (SXFEL)[46] to generate high intensity X-ray vortices with wavelengths of 1 nm. This configuration has two distinct advantages. First, the setup is very simple, since it only requires the addition of a few more helical undulators to the existing polarization control configuration. Second, it does not need special devices to produce and preserve the spiral bunched e beam structure during e beam transport, compared to previous methods.

2 Theory and modeling of higher harmonics from helical undulators

As mentioned earlier, X-ray photons carrying OAM can be generated in particle accelerators by converting an X-ray beam[18-23] or using helical undulators[39]. This triggers a careful re-examination of the angular momentum properties of emission from charged particles in circular and spiral motion[47-51].

We know that there are different kinds of undulator magnetic fields and possibly various structures of external lasers in diverse undulator devices. When the relativistic electron beam (e beam) is injected into the undulator, its energy and the eventual density distribution are affected by the magnetic and laser fields. In other words, the energy and final density distribution of the relativistic e beam are determined by the three-dimensional structure of the resonant interaction between the e beam, undulator field, and external injected laser. After the resonant interaction, the microbunched e beam is produced. These microbunched e beams can then be used to radiate electromagnetic waves in different downstream radiation facilities. Special light beams can be generated by resonant coupling with higher order harmonics. Therefore, it may be used to radiate fields with special 3D structures by properly tuning the injected microbunching and harmonic resonances in the downstream radiator. The helical light beams carrying OAM, generated in Ref.([42]), arise from the transmission of the structural information of the helical microbunched e beams to the optical field.

We use the standard linearized FEL equation to describe the interaction between the injected e beams and input laser in the planar undulator. After some discussion, the density perturbation n1(r) can be expressed as the sum of the expanded eigenmodes. The density modulation expansion is[52]:

n1(r)=kwϵ0ejaj(z)ε˜j(r), (1)

where kw is the wave number (the undulator wavelength is λw=2π/kw) and ϵ0 is the vacuum dielectric constant.

The evolution of the harmonic amplitudes can be expressed as[52]:

d2dz2aq(z)+θp2jFq,jaj(z)=qDq,q(h)Cq(z)eiθq(h)z, (2)

where θp is the relativistic longitudinal plasma wave number and aq(0), daq(0)/dz, and Cq(0) indicate the bunching, velocity modulation, and amplitude, respectively, of the incident field in the initial case. The second term on the left-hand side of equation(2) comes from the longitudinal space charge. The coefficient Dq,q(h) [52]on the right-hand side gives the coupling condition between the electron beams and harmonic fields.

Dq,q(h)=g^θp2cK[kzq(ω)+kw]22γωkwFq,q(h), (3)

where c is the speed of light, K is the undulator parameter, ω is the frequency, g^ denotes the electron motion, Fq,q(h) indicates the overlay effect of the e beam and incident laser lights, and f(r) represents the transverse density distribution of electron beams. After interacting with a typical harmonic field incident mode, the e beam modulation in equation (1) can be expressed as follows[52]:

ε˜q(r)=Rpl(r)eilϕ. (4)

In the above formula, the mode index q includes two coefficients, q=(p, l), where p and l correspond to the radial and azimuthal modes, respectively. When the e beam distribution is axisymmetric f(r)=f(r), the overlay effect of the e beam and incident fields can be straightforward[52]

F(p,l),(p,l)(h)2πδl,l±(h1). (5)

We can see the relationship between l, l’, h and the polarization direction of the undulator, ±, where l and l’ represent the radial modes about the azimuth of the electron beams and electromagnetic fields. There are some simple cases. For example, when an incident electromagnetic wave in a random l’ mode is injected into the undulator in the fundamental mode (h=1), the homologous azimuthal modes can be obtained in the electron beams l=l’. When an axisymmetric electromagnetic wave l’=0 is injected into the undulator working at the second harmonic mode (h=2), a spiral bunched electron beam is formed, where the helical undulator can be either right-handed (l=-1) or left-handed (l=1). Therefore, corkscrewed micro-bunched electron beams at the second harmonic can be produced by the interaction between an initially unmodulated electron beam and incident Gaussian electromagnetic field (laser field).

We now discuss the properties of emission from helical undulators. The time-dependent complex field amplitude A of radiation from a pure circular undulator is[39]:

A=(AxiAy)/2=2expi(n1)ϕ[(γθnKX)Jn(X)KJn(X)]. (6)

Here, X=2nξγθK, ξ=1/(1+γ2θ2+K2), K is the undulator deflecting parameter, θ is the viewing angle of the propagation axis, ϕ is the azimuth of the observer’s location in the transverse plane perpendicular to the direction of propagation, γ is the Lorentz factor, Jn is the nth order Bessel function of the first kind, and Jn(X) is the first derivative of Jn with respect to X. It is clear that the above equation is typical of Laguerre-Gaussian (LG) modes with a topological charge of l=n-1. LG beams are a group of solutions of the paraxial wave equation in the cylindrical coordinate system, and they are characterized by the OAM of a photon with lℏ. The LG modes are expressed by[3]:

uLG(r,θ,z)p,lIp,l(r,z)exp[i(2p+l+1)tanz/zR1]exp(ilθ). (7)

Similar to the above example, the last term on the right in equation (7) shows that there is an azimuthal component of the phase, where p is the radial index which decides the radial intensity distributions and l is the azimuthal index. The wavefront (or the phase structure) is a spiral surface when l≠0.

As an example, we perform a simulation using the current SSRF (Shanghai Synchrotron Radiation Facility)[53-55] beam parameters (3.5 GeV beam energy, 3 nm-rad horizontal emittance, and 1% coupling) and an actual magnetic field (K∼0.78) of the 32-mm-period helical undulator, whose total length is 4.5 m. The radiation properties of such helical undulators are simulated using the SPECTRA code[56].

The trajectory of an electron moving inside the undulator is a spiral along the axis in the direction of propagation. The radiation produced by the electron is emitted along the direction of propagation within a very small divergence angle. The spectral flux density of the light from the helical undulator is shown in Fig. 1, which indicates that the fundamental and second harmonic modes are at 2256 eV and about 4413 eV within two orders of magnitude, respectively. We then calculate the complex amplitude spatial profile of the radiation field on the cross section perpendicular to the direction of propagation, and estimate the phases of the first and second harmonic radiation using the real and imaginary parts of Ex and Ey, which are shown in Fig. 2. We observe that the transverse flux density profile of the fundamental radiation is Gaussian, while the transverse flux density distribution of the second harmonic is nearly annular. The typical transverse phase distribution of the second harmonic light is a spiral surface along the direction of propagation, which indicates that the photons carry an OAM (in this case, a topological charge l=-1).

Fig. 1.
The spectral flux density of the helical undulator.
pic
Fig. 2.
(Color online) The transverse intensity distribution and phase of the first (top row) and the second harmonic (bottom row) of the radiation from the helical undulator.
pic

It turns out that a relativistic e beam will move spirally along the axis of propagation after entering the helical undulator. Owing to this special motion mode, the light produced by the moving electrons forms a spiral structure at higher harmonics indicating that the photons carry OAM.

3 Theory of Self-Amplified Spontaneous Emission in Free-Electron Lasers

A revolutionary development in X-ray production was the progress of XFEL[57]. When an electron beam enters a periodic magnetic structure, the coherent radiation is amplified and this is the basis of the FEL. The radiation generated in an XFEL interacts with the beam of electrons in return to get more excellent X-ray than light sources based on storage rings in terms of intensity and coherence. At present, the high-energy e beams produced by a linac can be used to drive high-gain XFEL amplifiers using a long undulator. The gain of the FEL amplifier can be high enough to allow the initial incoherent undulator radiation to evolve into a high intensity quasi-coherent electromagnetic field——SASE[58].

In principle, an FEL can work at any wavelength, limited only by the energy and quality of the e beams obtained by the accelerator, and thus, can be used to fill gaps in the spectrum that cannot be covered by other coherent sources. At present, FEL oscillators based on electrostatic accelerators, cyclotrons, and rf-linear accelerators have been built. FEL oscillators in the storage rings are suitable for producing coherent radiation from visible to ultraviolet wavelengths[58]. The main technique for producing coherent radiation at shorter wavelengths involves a high gain FEL that does not require a reflector. In the late 1990s and early 2000s, devices were first designed and built to produce high-intensity ultraviolet and vacuum ultraviolet radiation. In 2005, the FLASH installed in Hamburg, Germany, began offering soft XFEL to scientific users[59]. In 2009, the Linac Coherent Light Source(LCLS) at SLAC ushered in the era of hard XFEL[60]. Subsequently, the XFEL devices in Japan(SACLA)[61] and Italy(FERMI)[71] were successfully debugged and provided services to users. At present, the soft XFEL device in Shanghai, China has been built and successfully tested[46]. We also mention the European XFEL in Germany[62], PAL XFEL at Posco in South Korea[63], and SwissFEL at PSI in Switzerland[64].

An FEL can operate in several modes like any laser, such as the FEL oscillator, amplifier, and SASE. The oscillator uses reflectors to constrain the radiation so that the radiation field passes through the undulator many times. An FEL can also function as a linear amplifier to amplify radiation with a central frequency close to the resonant condition. Owing to the need for seed light of the corresponding wavelength, this mode of operation is most easily implemented in the spectral region already covered by other types of light sources. FEL amplifiers extending to shorter wavelengths often require advanced techniques such as harmonic generation, so it is challenging to apply them to X-ray wavelengths. It is for this reason that the first scale and built XFEL devices were based on the SASE mode[58].

SASE originates from the growth and saturation of spontaneous emission in a long undulator, without the reflector and oscillator. Although the SASE radiation is not completely coherent, the gain process increases the brightness of the source to about 10 orders of magnitude higher than that of traditional synchrotron sources. In order to learn more about SASE, we consider the radiation field of the FEL process given by[58]

a(z^)13[a(0)ib(0)μ3iμ3P(0)]eiμ3z^. (8)

The equation(8) is approximately obtained considering a long propagation distance, where z^1. The first term in square brackets of the formula(8) represents the coherent amplification of the external radiation signal, while the second and third terms in square brackets represent the FEL radiation caused by density and energy modulation, respectively, of the electron beams. When these modulations occur due to the shot noises of e beams, the exponential growth is known as SASE. In equation (8), where the normalized ordinate is z^2kuρz, and ρ is the nondimensional Pierce parameter[65],

ρ=[e2K2[JJ]2ne32ϵ0γr3mc2ku2]1/3. (9)

Here, the factor [JJ] comes from the longitudinal oscillation of the electrons in the plane undulator, which alters the average coupling between the particle and radiation field[58].

[JJ]J0(K24+2K2)J1(K24+2K2). (10)

The parameters μ1, μ2, and μ3 [66]are

μ1=1,μ2=13i2,μ3=1+3i2. (11)

We then define three parameters to analyze the characteristics of SASE radiation[58].

dadz^=b (12) dbdz^=iP (13) dPdz^=a (14)

where a is the radiation field which represents the coherent radiation drives energy modulation, b is the cluster factor which reflects the cluster- generated coherent radiation, and P is the collective momentum which reflects the transformation from energy modulation to density modulation. The resonance condition of the undulator is

kuk=λλu=1+K2/22γ2. (15)

For convenience, we assume that there is no external field(a(0)=0) and the divergence of beam energy is 0 (P(0)=0). Thus, in the exponential growth range, the radiation field is[58]

|a(z^)|219|b(0)|2e3z^. (16)

Here, by normalizing the propagation distance 3z^=3(2kuzρ)=z/LG0, the ideal one-dimensional power gain length is

LG0λu4π3ρ. (17)

The cluster factor 〈|b(0)|2〉 at the entrance of the undulator comes from the initial shot noises of the e beams, which have been amplified during FEL. The level of shot noises is determined by the number of particles in the radiation coherence length. The cluster factor is[58],

|b(0)|2=1Nlcoh2|jlcoheiθj|21Nlcoh, (18)

where Nlcoh is the number of electrons in the scale of the coherence length, normalized bandwidth of SASE is Δω/ωρ, coherence time is tcohλ1/, and coherence length is lcohλ1/ρ. Thus, the coherence length can be regarded as the shift of the radiation relative to the electron beam over several gain lengths. Therefore, the starting shot noise of the SASE FEL is represented by the following equation:

Nlcoh1ecλ1ρ. (19)

We note that any SASE pulse is essentially an amplified undulator radiation. The one-dimensional power spectrum density can be used to calculate the radiated energy of the undulator[58]:

Uund=TdωdϕdPdωdϕTdωλ2AtrdPdωdϕ|ϕ=0 (20)

Here λ2/tr can be understood as the characteristic angular divergence of a source, whose area is tr. Therefore,

dPdω|1D=λ2Atrδ(ϕ)dPdω=λ2AtrdPdωdϕ|ϕ=0. (21)

Then, at the saturation length of the FEL (Nu≈1/ρ), the saturated FEL energy can be rewritten as[58]:

UFEL=Neργrmc2=Neρω1TUund8π (22) tcohNeTUund=NlcohUund=TtcohNlcoh2UundNe. (23)

The equation (22) indicates that the output of the saturated FEL is much stronger than the undulator emissions. Their ratio is the number(Nlcoh>105) of particles in the coherence time. The energy of the FEL is the product of the single electron radiation energy of the undulator, square of the number of electrons in the coherence length and number of coherent regions T/tcoh.

In conclusion, the SASE FEL emits radiation from shot noise and the electron beams in the undulator can be distributed in the scale of the radiation wavelength to form periodic density modulations and achieve coherent enhancement of radiation intensity. The density modulation originates from the resonant interaction between the electron beams and X-rays in the periodic undulator. Hence, when the undulator is long enough, current of the electron beams is sufficiently large, and electron beams are of good quality(low emissivity and low energy divergence), the electrons will generate periodic density modulation at the scale of the resonant X-ray wavelength after interacting with the radiation. This results in intensities higher than incoherent undulator radiation, even though the electron beam is much longer than the radiation wavelength. In this work, we use the SASE FEL to obtain periodic microbunching e beams for generating high-quality XOAM.

4 Scheme for generating coherent X-rays carrying OAM at the SXFEL

As indicated in the recent study of the interaction between atoms and XUV spiral light[67], it is necessary to observe the non-dipole transition optionally when the photon beam has a single clear phase singularity or the position of the sample atom is close to the phase singularity. It is suggested[67] that the diffraction-limited storage ring[68] generates such XUV or X-ray spiral light under special conditions and an FEL could produce coherent vortex beams taking advantage of diverse schemes.

The phased project SXFEL[46] mainly includes the SXFEL test facility (SXFEL-TF) and SXFEL user facility (SXFEL-UF). The SXFEL-TF is mainly used to promote the research of FEL in China, such as investigating the feasibility of seed XFEL with two-stage cascade HGHG-HGHG, proposing a brand-new plan based on an EEHG-HGHG cascade, and promoting the research and development of critical XFEL technology. Finally, the Chinese Academy of Sciences decided to build the SXFEL-TF at the SSRF in Shanghai. The initial commissioning was successfully performed from 2014 until the beginning of 2020. The Shanghai Institute of Applied Physics (SINAP) cooperated with the Shanghai-Tech University to upgrade the SXFEL-TF to a water window SXFEL-UF, including the development of the scientific experimental line station, upgradation of the energy of the linear accelerator to 1.5 GeV, construction of the second SASE FEL facility, as well as the construction of other necessary public facilities. It is expected to open to the public user by the end of this year.

In the newly added second SASE FEL branch, a typical polarization control setup (using helical undulators as afterburners) is being planned. We use a configuration with minor modifications to the baseline design to generate X-rays carrying OAM, as shown in Fig. 3. This special method is analyzed in greater detail in ref[69]. The first SASE undulators are used for generating microbunching from the electron beam, and the microbunched electron beams are then sent to the downstream helical undulators to generate coherent X-rays carrying OAM with wavelengths of ∼1 nm at the second harmonic of the helical undulators. The main parameters are summarized in Table 1.

Fig. 3.
(Color online) The propsed FEL configuration for generating X-rays carrying OAM at the SXFEL.
pic
Table 1.
Main parameters of the X-rays carrying OAM at the SXFEL.
Parameters Value
e- beam energy (GeV) 1.5
e- energy spread 0.01%
e- bunch charge (nC) 0.5
e- normalized emittance(mmmrad) 0.6
e- pulse length (FWHM) (μm) 50
e- repetition rate (Hz) 10
SASE undulator K 1.6
SASE undulator period (mm) 15
SASE undulator length/segment (m) 4
SASE undulator No.of segment 4
Helical undulator K 2.3
Helical undulator period (mm) 15
Helical undulator length/segment (m) 4
Helical undulator No.of segment 2
Show more

We use the SIMPLEX code[70] to simulate this process with time dependence. First, we simulated the SASE FEL part with six planar undulators, and the gain curve of the first harmonic of FEL radiation at λ1=2 nm is shown in Fig. 4(left). In order to achieve better FEL performance downstream, we selected four SASE undulator segments owing to the best total bunching factor at z=18.9 m (shown in Fig. 4(Right)) and relatively lower energy spread, compared to the saturation point.

Fig. 4.
(a) Gain curve of the SASE FEL with 6 SASE undulators;(b) Bunching factor of the electron beam at the end of the 4th SASE undulator (z=18.9m)
pic

In our simulation, the focusing magnets are periodically arranged to be compatible with the layout of the undulator. We select the FUDU(QF-u-QD-U) lattice type to satisfy our requirements. Since the gain of the FEL is related to the e beam size and angular spread, there is an optimum betatron function that maximizes the gain. Thus, the lattice function should be selected to realize the optimum betatron function over the whole undulator line. The field gradient of the focusing quadrupole magnets in the upstream planar undulator segments is 6.50744 T/m and the field gradient of the defocusing quadrupole magnets is -3.50995 T/m. The focusing and defocusing quadrupole magnets are both 0.2 m long. In addition to the lattice type, the initial Twiss parameters at the entrance of the undulator are listed as follows. The horizontal (x) and vertical (y) betatron functions are 11.8631 m and 4.61501 m, respectively. while the horizontal (x) and vertical (y) alpha functions(corresponding to the derivative of the betatron function) are 1.59554 and -0.68688, respectively. The horizontal(x) and vertical(y) betatron functions along all the four SASE undulator segments are shown in Fig. 5(left). Moreover, the lattice type in the afterburner helical undulator segments is FUDU(QF-U-QD-U), and the field gradients of the focusing and defocusing quadrupole magnets are 4.26216 T/m and -4.27089 T/m, respectively. The focusing and defocusing quadrupole magnets are both 0.2 m long. The horizontal (x) and vertical (y) betatron functions are 11.462 m and 5.04241 m, respectively, while the horizontal (x) and vertical (y) alpha functions are 0.966439 and -0.520785. The horizontal (x) and vertical (y) betatron functions along the afterburner are shown in Fig. 5(right).

Fig. 5.
(Color online) The betatron function of the whole undulator segments.(a) The betatron function along the whole 4 SASE undulator segments; (b) The betatron function of the afterburner (2 helical undulator segments).
pic

In the SASE FEL, there is no special distribution of e beams when they first enter the undulator. However, the unstable SASE FEL makes the electrons produce shot noises which act on the electrons after amplification and redistribute the e beams. The profile of the emission from the device is determined by the microbunching structure and geometric configuration of the FEL radiator. We know that the emission at the fundamental modes of both planar and helical undulators has a strong peak on the axis and the screw-type phase of the XOAM originates from the spiral microbunched e beams. Based on the above discussions, these fine-tuned microbunched e beams require precise control over the electron profile upstream of the circularly polarized undulator with a particular input laser field. The fine-tuned microbunched e beam is then injected into the downstream radiator to emit and amplify coherent XOAM light. Thus, the above method requires the careful preservation of the spiral structure of the microbunching e beams when propagating into the downstream undulator.

While it is relatively difficult to maintain the spiral structures of the microbunched e beams, our simulated method does not require the preservation of the spiral structures. High-quality periodic micro-bunched electron beams are obtained with the FUDU lattice design. The micro-bunched e beams are then injected into the downstream helical undulator. The helical undulator has the same periodic structural parameters as the upstream planar undulator to obtain X-ray vortices at the second harmonic at the wavelength λ2=1 nm. The gain curve along z, transverse intensity, and horizontal phase at the exit z=8.9 m are shown in Fig. 6. The typical donut shaped spatial distribution of the intensity distribution and spiral phase indicate that the second harmonic carries OAM with an azimuthal mode index of l=-1. The X-ray OAM intensity is at the sub-μJ level. We use the SASE FEL to produce microbunched e beams with a specific periodic structure when the preservation of the spiral microbunched e beams is unnessesary. Hence, this is an easier method for generating high energy XOAM.

Fig. 6.
(Color online) (a) Gain curve of the 2nd harmonic of the helical afterburner undulators; (b) Transverse intensity distribution of the 2nd harmonic of FEL radiation from the helical afterburner undulators at the exit; (c) Phase of the 2nd harmonic of FEL radiation from the helical afterburner undulators at the exit
pic

We also discuss the stability of the pulse energy of the second harmonic radiation using the same parameters for the SASE and helical undulator but different initial macroparticle distributions, i.e., random numbers auto seeding the SASE FEL,under the time-dependent state simulation mode. In the first section of the undulator, spontaneous radiation and periodic micro-bunched e beams are generated, which then act as source e beams for the X-ray carrying OAM generation process. After ten simulations with various initial shot noise seeding in the SASE section, we find that the evolution of the bunching factor at the first harmonic along the direction of propagation in the planar undulator and the pulse energy of the second harmonic radiation in the helical undulator are similar, and stable growth within the error range. These are shown in Fig. 7. Therefore, using the SASE FEL mode as the upstream setup, it is possible to generate similar well-bunched microbunching e beams, despite fluctuations of the initial macroparticle distribution producing shot noise seeding in the SASE. The pulse energy of the second harmonic of the helical undulator is relatively stable, guaranteeing the stability of this method for generating XOAM.

Fig. 7.
(Color online) (a) the bunch factors in the same 4 SASE undulator segments with 10 different shot noise seedings in SASE undulator; (b) the pulse energy of the 2nd harmonic radiation in the same 2 helical undulator segments with 10 different seedings in SASE undulator.
pic

It is known that when microbunched relativistic electron beams pass through the helical undulator and produce resonant interactions, the axisymmetric incident laser field transmits the structural information to the electron beams. Thus, an electron beam propagating through a long helical undulator can generate many microbunches of electrons. Each microbunch of electron beams is a helix, which can produce OAM radiation in harmonics, whose radiation powers are relatively weak. To increase the intensity of the vortex light, the special helical microbunched electron beams could be regarded as sources of OAM emission in a radiator, while carefully maintaining the spiral density distribution during propagation in the downstream undulator. In our simulation, the well-defined microbunches of electron beams are generated in the upstream planar undulator, which are then injected directly into the downstream short helical undulator. This is tuned to the harmonic and thus, the bunched e beam radiates coherently in the afterburner. The resultant vortex radiation carrying OAM is generated on the harmonics. In the downstream afterburner, only the parameter K of the helical undulator is changed. This enables it to emit coherent radiation at the second harmonic of the helical undulator, but owing to the weaker interaction between the microbunched e beam and helical undulator radiation at the second harmonic, the intensity of the OAM radiation emitted at the second harmonic is smaller than that at the first harmonic.

In order to compare with the FERMI-like but different all-helical-undulator configuration[71], which is not externally seeded possesses the same set of parameters for the accelerator beam and undulators, and only differing in the use of a helical undulator rather than linear one, we simulate its FEL output at the second harmonic. It is shown that in the afterburner configuration (Fig. 3), the FEL intensity quickly saturates within 10 m, while in the FERMI-like all-helical-undulator configuration, the FEL intensity grows slowly to saturation in 25 m, with similar FEL intensity output. However, if we use the exact FERMI configuration with external seeding, the afterburner type setup can be easily mimicked to obtain fully coherent OAM. To summarize, there are two advantages of this simple afterburner setup using the existing devices in SXFEL at Shanghai. First, the short helical undulators can be directly added after the existing SASE planar undulators, without the need to generate and preserve the complicated helical electron distribution. Second, this afterburner configuration is more cost-effective and efficient since it is based on existing devices and there is no need to build a new all-helical undulator.

5 Conclusion

We simulate a straightforward and easy-to-implement helical undulator afterburner configuration at the SXFEL to generate high power X-rays carrying OAM with wavelengths of 1 nm. We use four SASE segments from the SASE FEL line at the SXFEL to produce periodic microbunched e beams. These are a consequence of the interaction between the initial randomly distributed electron beams and shot noise in the SASE process. The microbunched e beams are then injected into the downstream helical undulator (two segments). The scheme does not need additional dispersion devices along the transport channel of the electron beams, which can then be used to obtain optical spiral radiation with topological charge l=-1. The minus sign of the topological charge is due to the right-handed helicity of the circularly polarized undulator. We select the FUDU(QF-u-QD-U) lattice type in the SASE planar undulator and downstream helical undulator to satisfy our requirements. Using different shot noises in the initial SASE undulator, the relatively stable results are obtained. After simulation, the short helical undulators can be added directly after the existing SASE planar undulator. This configuration is more cost-effective and efficient as it is based on existing devices.

As far as we know, this is the first demonstration of the generation of intense X-ray vortices whose wavelength is ∼1 nm at the SXFEL in China. We expect that exciting experiments with OAM carrying X-rays at the SXFEL will be realized as soon as the SASE facility is completed and the structure of the helical undulator is well processed. These X-rays with OAM property in the "magnetic window" can be used in various fields of research, especially for exploring magnetic dichroism [34] and detecting microscopic magnetic vortices[35].

References
[1] H. Rubinsztein-Dunlop, A. Forbes, M. V. Berry et al.,

Roadmap on structured light

. Journal of Optics 19, 013001(2016). https://doi.org/10.1088/2040-8978/19/1/013001
Baidu ScholarGoogle Scholar
[2] J. F. Nye, M. V. Berry,

Dislocations in wave trains

. Proc. R. Soc. Lond. A 336, 0012(1974). https://doi.org/10.1098/rspa.1974.0012
Baidu ScholarGoogle Scholar
[3] L. Allen, M. W. Beijersbergen, R. J. C. Spreeuw, J. P. Woerdman,

Orbital angular momentum of light and the transformation of Laguerre-Gaussian laser modes

. Phys.Rev. A 45, 8185(1992). https://doi.org/10.1103/PhysRevA.45.8185
Baidu ScholarGoogle Scholar
[4] L. Allen, S. M. Barnett, M. J. Padgett (Eds.), Optical Angular Momentum (Institute of Physics Pub., Bristol, Philadelphia, 2003).
[5] G. Molina-Terriza, J. P. Torres, L. Torner,

Twisted photons

. Nat. Phys. 3, 305(2007). https://doi.org/10.1038/nphys607
Baidu ScholarGoogle Scholar
[6] S. Franke-Arnold, L. Allen, M. Padgett,

Advances in optical angular momentum

. Laser Photonics Rev. 2, 299(2008). https://doi.org/10.1002/lpor.200810007
Baidu ScholarGoogle Scholar
[7] A. M. Yao, M. J. Padgett,

Orbital angular momentum: origins, behavior and applications

. Adv. Opt. Photonics 3, 161(2011). https://doi.org/10.1364/AOP.3.000161
Baidu ScholarGoogle Scholar
[8] J. P. Torres, L. Torner (Eds.), Twisted photon: Applications of Light with Orbital Angular Momentum. 2011 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim.
[9] D. L. Andrews, M. Babiker (Eds.), The Angular Momentum of Light, Cambridge University Press 2013.
[10] R. W. Bowman, M. J. Padgett,

Optical trapping and binding

. Rep. Prog. Phys. 76, 026401(2013). https://doi.org/10.1088/0034-4885/76/2/026401
Baidu ScholarGoogle Scholar
[11] M. J. Padgett,

Orbital angular momentum 25 years on

. Opt. Express 25, 11265(2017). https://doi.org/10.1364/OE.25.011265
Baidu ScholarGoogle Scholar
[12] Y.J. Shen, X.J. Wang, Z.W. Xie et al.,

Optical vortices 30 years on: OAM manipulation from topological charge to multiple singularities

. Light: Science & Applications 8, 90(2019). https://doi.org/10.1038/s41377-019-0194-2
Baidu ScholarGoogle Scholar
[13] D. S. Bykov, S.R. Xie, R. Zeltner et al.,

Long-range optical trapping and binding of microparticles in hollow-core photonic crystal fibre

. Light: Science & Applications 7, 22(2018). https://doi.org/10.1038/s41377-018-0015-z
Baidu ScholarGoogle Scholar
[14] N.S. Huang, K. Li, H.X. Deng,

BRIGHT: the three-dimensional X-ray crystal Bragg diffraction code

. Nucl. Sci. Tech. 30, 39(2019). https://doi.org/10.1007/s41365-019-0559-5
Baidu ScholarGoogle Scholar
[15] H.L. Wu, J.H. Chen, Z. T. Zhao et al.,

Proposal for a high brightness γ-ray source at the SXFEL

. Nucl. Sci. Tech. 26, 050103(2015). https://doi.org/10.13538/j.1001-8042/nst.26.050103
Baidu ScholarGoogle Scholar
[16] A. G. Peele, P. J. McMahon, D. Paterson et al.,

Observation of an x-ray vortex

. Opt. Lett. 27, 1752(2002). https://doi.org/10.1364/OL.27.001752
Baidu ScholarGoogle Scholar
[17] A. G. Peele, K. A. Nugent, A. P. Mancuso et al.,

X-ray phase vortices: theory and experiment

. J. Opt. Soc. Am. A 21, 1575(2004). https://doi.org/10.1364/JOSAA.21.001575
Baidu ScholarGoogle Scholar
[18] B. Terhalle, A. Langner, V. A. Guzenko et al.,

Generation of extreme ultraviolet vortex beams using computer generated holograms

. Optics Letters 36, 4143(2011). https://doi.org/10.1364/OL.36.004143
Baidu ScholarGoogle Scholar
[19] A. Sakdinawat, Y.W. Liu,

Soft-x-ray microscopy using spiral zone plates

. Optics Letters 32, 2635(2007). https://doi.org/10.1364/OL.32.002635
Baidu ScholarGoogle Scholar
[20] D. Cojoc, B. Kaulich, A. Carpentiero et al.,

X-ray vortices with high topological charge

. Microelectronic Engineering 83, 1360(2006). https://doi.org/10.1016/j.mee.2006.01.066
Baidu ScholarGoogle Scholar
[21] Y. Kohmura, K. Sawada, M. Taguchi et al.,

Formation of x-ray vortex dipoles using a single diffraction pattern and direct phase measurement using interferometry

. Appl. Phys. Lett. 94, 101112(2009). https://doi.org/10.1063/1.3095828
Baidu ScholarGoogle Scholar
[22] J. C. T Lee, S. J. Alexander, S. D. Kevan et al.,

Laguerre-Gauss and Hermite-Gauss soft X-ray states generated using diffractive optics

. Nat. Photon. 13, 205(2019). https://doi.org/10.1038/s41566-018-0328-8
Baidu ScholarGoogle Scholar
[23] L. Loetgering, M. Baluktsian, K. Keskinbora et al.,

Generation and characterization of focused helical x-ray beams

. Sci. Adv. 6, eaax8836(2020). https://dx.doi.org/10.1126/sciadv.aax8836
Baidu ScholarGoogle Scholar
[24] D. Seipt, A. Surzhykov, S. Fritzsche,

Structured x-ray beams from twisted electrons by inverse Compton scattering of laser light

. Phys. Rev. A 90, 012118(2014). https://dx.doi.org/10.1103/PhysRevA.90.012118
Baidu ScholarGoogle Scholar
[25] V. Petrillo, G. Dattoli, I. Drebot et al.,

Compton Scattered X-Gamma Rays with Orbital Momentum

. Phys. Rev. Lett. 117, 123903(2016). https://journals.aps.org/prl/abstract/10.1103/PhysRevLett.117.123903
Baidu ScholarGoogle Scholar
[26] M. Zurch, C. Kern, P. Hansinger et al.,

Strong-field physics with singular light beams

. Nature physics 8, 743(2012). https://doi.org/10.1038/nphys2397
Baidu ScholarGoogle Scholar
[27] G. Gariepy, J. Leach, K. T. Kim et al.,

Creating High-Harmonic Beams with Controlled Orbital Angular Momentum

. Phys. Rev. Lett. 113, 153901(2014). https://dx.doi.org/10.1103/PhysRevLett.113.153901
Baidu ScholarGoogle Scholar
[28] C. Hernandez-Garcia, J. Vieira, J. T. Mendonca et al.,

Generation and Applications of Extreme-Ultraviolet Vortices

. Photonics 4, 28(2017). https://doi.org/10.3390/photonics4020028
Baidu ScholarGoogle Scholar
[29] L. Rego, K. M. Dorney, N. J. Brooks et al.,

Generation of extreme-ultraviolet beams with time-varying orbital angular momentum

. Science 364, eaaw9486(2019). https://doi.org/10.1126/science.aaw9486
Baidu ScholarGoogle Scholar
[30] J. Vieira, R.M.G.M. Trines, E.P. Alves et al.,

Amplification and generation of ultra-intense twisted laser pulses via stimulated Raman scattering

. Nature Communications 7, 10371(2016). https://doi.org/10.1038/ncomms10371
Baidu ScholarGoogle Scholar
[31] J. Vieira, R. M. G. M. Trines, E. P. Alves et al.,

High Orbital Angular Momentum Harmonic Generation

. Phys. Rev. Lett. 117, 265001(2016). https://doi.org/10.1103/PhysRevLett.117.265001
Baidu ScholarGoogle Scholar
[32] A. Picon, A. Benseny, J. Mompart et al.,

Transferring orbital and spin angular momenta of light to atoms

. New Journal of Physics 12, 083053(2010). https://doi.org/10.1088/1367-2630/12/8/083053
Baidu ScholarGoogle Scholar
[33] A. Picon, J. Mompart, J. R. V. de Aldana, et al.,

Photoionization with orbital angular momentum beams

. Opt. Express 18, 3660(2010). https://doi.org/10.1364/OE.18.003660
Baidu ScholarGoogle Scholar
[34] M. v. Veenendaal, I. McNulty,

Prediction of Strong Dichroism Induced by X Rays Carrying Orbital Momentum

. Phys. Rev. Lett. 98, 157401(2007). https://link.aps.org/doi/10.1103/PhysRevLett.98.157401
Baidu ScholarGoogle Scholar
[35] M. v. Veenendaal,

Interaction between x-ray and magnetic vortices

. Phys. Rev. B 92, 245116(2015). https://doi.org/10.1103/PhysRevB.92.245116
Baidu ScholarGoogle Scholar
[36] G. Friesecke, R. D. James, D. Juestel,

Twisted X-Rays: Incoming Waveforms Yielding Discrete Diffraction Patterns for Helical Structures

. SIAM J. Appl. Math. 76, 1191(2016). https://doi.org/10.1137/15M1043418
Baidu ScholarGoogle Scholar
[37] D. Juestel, G. Friesecke, R. D. James,

Bragg-von Laue diffraction generalized to twisted X-rays

. Acta Cryst. A 72, 190(2016). https://dx.doi.org/10.1107/S2053273315024390
Baidu ScholarGoogle Scholar
[38] L. Ye, J. R. Rouxel, S. Asban et al.,

Probing Molecular Chirality by Orbital-Angular-Momentum-Carrying X-ray Pulses

. J. Chem. Theory Comput. 15, 4180(2019). https://doi.org/10.1021/acs.jctc.9b00346
Baidu ScholarGoogle Scholar
[39] S. Sasaki, I. McNulty,

Proposal for Generating Brilliant X-Ray Beams Carrying Orbital Angular Momentum

. Phys. Rev. Lett. 100, 124801(2008). https://doi.org/10.1103/PhysRevLett.100.124801
Baidu ScholarGoogle Scholar
[40] J. Bahrdt, K. Holldack, P. Kuske et al.,

First Observation of Photons Carrying Orbital Angular Momentum in Undulator Radiation

. Phys. Rev. Lett. 111, 034801(2013). https://doi.org/10.1103/PhysRevLett.111.034801
Baidu ScholarGoogle Scholar
[41] T. Kaneyasu, Y. Hikosaka, M. Fujimoto et al.,

Observation of an optical vortex beam from a helical undulator in the XUV region

. J. Synchrotron Rad. 24, 934(2017). https://doi.org/10.1107/S1600577517009626
Baidu ScholarGoogle Scholar
[42] E. Hemsing, A. Knyazik, M. Dunning et al.,

Coherent optical vortices from relativistic electron beams

. Nat. Phys. 9, 549(2013). https://doi.org/10.1038/nphys2712
Baidu ScholarGoogle Scholar
[43] E. Hemsing, M. Dunning, C. Hast et al.,

First Characterization of Coherent Optical Vortices from Harmonic Undulator Radiation

. Phys. Rev. Lett. 113, 134803(2014). https://doi.org/10.1103/PhysRevLett.113.134803
Baidu ScholarGoogle Scholar
[44] P. R. Ribic, D. Gauthier, G. D. Ninno,

Generation of Coherent Extreme-Ultraviolet Radiation Carrying Orbital Angular Momentum

. Phys. Rev. Lett. 112, 203602(2014). https://doi.org/10.1103/PhysRevLett.112.203602
Baidu ScholarGoogle Scholar
[45] P. R. Ribic, B. Rosner, D. Gauthier et al.,

Extreme-Ultraviolet Vortices from a Free-Electron Laser

. Phys. Rev. X 7, 031036(2017). https://doi.org/10.1103/PhysRevX.7.031036
Baidu ScholarGoogle Scholar
[46] Z.T. Zhao, D. Wang, Q. Gu et al.,

Status of the SXFEL Facility

. Appl. Sci. 7, 607(2017). https://doi.org/10.3390/app7060607
Baidu ScholarGoogle Scholar
[47] A. Afanasev, A. Mikhailichenko,

On generation of photons carrying orbital angular momentum in the helical undulator

. arXiv:1109.1603(2011).
Baidu ScholarGoogle Scholar
[48] M. Katoh, M. Fujimoto, H. Kawaguchi et al.,

Angular Momentum of Twisted Radiation from an Electron in Spiral Motion

. Phys. Rev. Lett. 118, 094801(2017). https://doi.org/10.1103/PhysRevLett.118.094801
Baidu ScholarGoogle Scholar
[49] M. Katoh, M. Fujimoto, N. S. Mirian et al.,

Helical Phase Structure of Radiation from an Electron in Circular Motion

. Sci. Rep. 7, 6130(2017). https://doi.org/10.1038/s41598-017-06442-2
Baidu ScholarGoogle Scholar
[50] H. Kawaguchi, M. Katoh,

Orbital angular momentum of Lienard-Wiechert fields

. Prog. Theor. Exp. Phys. 2019, 083A02(2019). https://doi.org/10.1093/ptep/ptz059
Baidu ScholarGoogle Scholar
[51] V. Epp, U. Guselnikova,

Angular momentum of radiation from a charge in circular and spiral motion

. Physics Letters A 383, 2668(2019). https://doi.org/10.1016/j.physleta.2019.05.038
Baidu ScholarGoogle Scholar
[52] E. Hemsing, P. Musumeci, S. Reiche et al.,

Helical electron-beam microbunching by harmonic coupling in a helical undulator

. Physical Review Letters 102, 174801(2009). https://doi.org/10.1103/PhysRevLett.102.174801
Baidu ScholarGoogle Scholar
[53] M.H.J , X. Yang, H.J. Xu et al.,

Shanghai Synchrotron Radiation Facility

. Chinese Science Bulletin 54, 4171(2009). https://doi.org/10.1007/s11434-009-0689-y
Baidu ScholarGoogle Scholar
[54] Z.C. Chen, Y.B. Leng, R.X. Yuan et al.,

Study of algorithms of phase advance measurement between BPMs and its application in SSRF

. Nuclear Science and Techniques 24, 010102(2013). https://doi.org/10.13538/j.1001-8042/nst.2013.01.007
Baidu ScholarGoogle Scholar
[55] Z.C. Chen, Y.B. Leng, R.X. Yuan et al.,

Beam position monitor troubleshooting by using principal component analysis in Shanghai Synchrotron Radiation Facility

. Nuclear Science and Techniques 25, 020102(2014). https://doi.org/10.13538/j.1001-8042/nst.25.020102
Baidu ScholarGoogle Scholar
[56] T. Tanaka, H. Kitamura,

SPECTRA: a synchrotron radiation calculation code

. Journal of Synchrotron Radiation 8, 1221(2001). https://doi.org/10.1107/S090904950101425X
Baidu ScholarGoogle Scholar
[57] C. Feng, H.X. Deng,

Review of fully coherent free-electron lasers

. Nuclear Science and Techniques 29, 160(2018). https://doi.org/10.1007/s41365-018-0490-1
Baidu ScholarGoogle Scholar
[58] K.-J. Kim, Z.R. Huang, R. Lindberg, Synchrotron Radiation and Free-Electron Lasers: Principles of Coherent X-Ray Generation. Cambridge University Press 2017.
[59] W. Ackermann, G. Asova, K. Zapfe et al.,

Operation of a free-electron laser from the extreme ultraviolet to the water window

. Nature Photonics 1, 336(2007). https://doi.org/10.1038/nphoton.2007.76
Baidu ScholarGoogle Scholar
[60] J. Galayda, J. Arthur, P. Anfinrud et al.,

Linac Coherent Light Source (LCLS) conceptual design report. SLAC, Report SLAC-R-593

, 2002
Baidu ScholarGoogle Scholar
[61] T. Shintake et al.,

sPring-8 Compact SASE Source Conceptual Design Report

. 2005. DOI: http://www-xfel.spring8.or.jp
Baidu ScholarGoogle Scholar
[62] R. Brinkmann, et al.

TESLA XFEL: First stage of the X-ray laser laboratory (Technical Design Report, Supplement). DESY, Report TESLA FEL 2002-09

, 2002.
Baidu ScholarGoogle Scholar
[63] H.-S. Kang, K.-W. Kim, I.-S. Ko,

Current status of PAL-XFEL project

. Proceedings of IPAC 2014, p. 2897, 2014
Baidu ScholarGoogle Scholar
[64] R. Ganter, et al.

SwissFEL-conceptual design report

. Paul Scherrer Institute (PSI), Report, 2010
Baidu ScholarGoogle Scholar
[65] R. Bonifacio, C. Pellegrini, L. M. Narducci,

Collective instabilities and high-gain regime in a free electron laser

. Opt. Commun. 50, 373(1984). https://doi.org/10.1016/0030-4018(84)90105-6
Baidu ScholarGoogle Scholar
[66] N. Kroll, W. McMullin,

Stimulated emission from relativistic electrons passing through a spatially periodic transverse magnetic field

. Phys. Rev. A, 17, 300(1978). https://doi.org/10.1103/PhysRevA.17.300
Baidu ScholarGoogle Scholar
[67] T. Kaneyasu, Y. Hikosaka, M. Fujimoto et al.,

Limitations in photoionization of helium by an extreme ultraviolet optical vortex

. Phys. Rev. A 95, 023413(2017). https://doi.org/10.1103/PhysRevA.95.023413
Baidu ScholarGoogle Scholar
[68] M. Eriksson, J. F. van der Veen, C. Quitmann,

Diffraction-limited storage rings - a window to the science of tomorrow

. J. Synchrotron Radiat. 21, 837(2014). https://doi.org/10.1107/S1600577514019286
Baidu ScholarGoogle Scholar
[69] E. Hemsing,

Coherent photons with angular momentum in a helical afterburner

. Physical Review Accelerators and Beams 23, 020703(2020). https://doi.org/10.1103/PhysRevAccelBeams.23.020703
Baidu ScholarGoogle Scholar
[70] T. Tanaka,

SIMPLEX: simulator and postprocessor for free-electron laser experiments

. Journal of Synchrotron Radiation 22, 1319(2015). https://doi.org/10.1107/S1600577515012850
Baidu ScholarGoogle Scholar
[71] C. J. Bocchetta, et al.

FERMI@Elettra CDR

, http://www.elettra.trieste.it/FERMI, (2007).
Baidu ScholarGoogle Scholar