High-brilliance, high-flux compact inverse Compton light source
aa r X i v : . [ phy s i c s . acc - ph ] M a r High-brilliance, high-flux compact inverse Compton light source
K. E. Deitrick, ∗ G. A. Krafft,
1, 2
B. Terzi´c, and J. R. Delayen
1, 2 Department of Physics, Center for Accelerator Science,Old Dominion University, Norfolk, Virginia 23529, USA Thomas Jefferson National Accelerator Facility, Newport News, Virginia 23606, USA (Dated: March 29, 2018)A compact Inverse Compton Light Source (ICLS) design is presented, with flux and brillianceorders of magnitude beyond conventional laboratory-scale sources and other compact ICLS designs.This design utilizes the physics of inverse Compton scattering of an extremely low emittance electronbeam by a laser pulse of rms length of approximately two-thirds of a picosecond (2/3 ps). Theaccelerator is composed of a superconducting radiofrequency (SRF) reentrant gun followed by fourdouble-spoke SRF cavities. After the linac are three quadrupole magnets to focus the electron beamto the interaction point (IP). The distance from cathode surface to IP is less than 6 meters, withthe cathode producing electron bunches with a bunch charge of 10 pC and a few picoseconds inlength. The incident laser has 1 MW circulating power, a 1 micron wavelength, and a spot size of3.2 microns at the IP. The repetition rate of this source is 100 MHz, in order to achieve a high fluxdespite the low bunch charge. The anticipated X-ray source parameters include an energy of 12 keV,with a total flux of 1 . × ph/s, the flux into a 0.1% bandwidth of 2 . × ph/(s-0.1%BW),and the average brilliance of 2 . × ph/(s-mm -mrad -0.1%BW). I. INTRODUCTION
Since their discovery in 1895, X-rays have been a pow-erful technique for determining the structure of con-densed matter. For the first 70 years of using X-rays,sources barely changed from the original bremsstrahlungtubes used in their discovery [1]. Until recently, largeaccelerator-based synchrotron facilities set the standardfor the highest quality X-ray beams. At present, thisstandard has been largely surpassed in free electron lasers(FELs).Most high-brilliance sources exist at large facilities,especially third-generation synchrotrons [2]. However,due to various concerns, among them cost, risk of trans-porting valuable items, and limited available runtime atlarge facilities, there has been an increasing demand forlaboratory-scale sources. Sometimes referred to as “com-pact”, one description is any machine that fits in a 100m area. Additional desirable constraints are that thepurchase and operating cost are not prohibitive for thesmaller facilities and that the operation of a such a ma-chine is possible by non-experts.There are many X-ray experimental techniques thatexist today; any given technique may be utilized in awide range of fields. Some of the more prominent tech-niques currently in use include phase contrast imaging(PCI), absorption radiography, K-edge subtraction imag-ing, radiotherapy (treatment of tumors with X-rays), andcomputed tomography (CT). Some of the fields in whichthese techniques are used include medicine, cultural her-itage, material science development, and industry [3, 4].Given the wide range of applications, the increasing de-mand for higher quality X-ray sources is understandable. ∗ [email protected]; now at Jefferson Lab; [email protected] FIG. 1. A schematic of the entire design. The first cryomod-ule contains the gun and two double-spoke cavities, the secondcontains the last two double-spoke cavities. Three quadrupolemagnets (red) follow the linac, before the interaction point(yellow).
In this paper, we present a design of a compact InverseCompton Source based on SRF beam acceleration whichwas outlined in [5]. Because the SRF is run continuouswave (CW), high average flux and brilliance are possible.This paper is organized as follows. A brief overview ofrelevant SRF, electron beam, and X-ray beam parame-ters is presented in Sec. II. Sec. III goes into detail aboutInverse Compton Light Sources, inverse Compton scat-tering, and compact ICLS designs. Our design consists ofthree separate regions: the SRF reentrant gun (Sec. IV),the SRF linear accelerator (Sec. V), and the final focusing(Sec. VI). A complete layout of these design componentsis shown in Fig. 1. In Sec. VII the sensitivity study re-sults are presented, while the incident laser is addressedin Sec. VIII. The complete X-ray beam parameters arepresented in Sec. IX, plans for future work are given inSec. X, before the final summary presented in Sec. XI.The design presented in this paper potentially outper-forms all other compact ICLS designs.
II. BEAM PARAMETERSA. Electron Beam Parameters
In our simulations, each particle in the beam is de-scribed by a set of six coordinates: ( x , p x , y , p y , z , p z ) where x and y are the transverse positions of theparticle, p x and p y are the transverse momenta, z isthe longitudinal position relative to a reference parti-cle, and p z is the momentum along the beam trajec-tory. For a free particle, the energy E of any particlewithin the bunch is related to its total momentum p by βE = cp = c q p x + p y + p z .Following standard practices, it is often more conve-nient to use an alternate set of coordinates: ( x , x ′ , y , y ′ , z , δ ) where x ′ ≡ p x /p z , y ′ ≡ p y /p z , and δ ≡ ∆ p/p suchthat ∆ p ≡ p − p , with p representing the momentum ofa particle with the average energy of the bunch. Whenthe relative momentum error δ is small, x ′ ≈ p x / h p z i .In this paper, beam sizes are quoted using the rootmean square ( rms ) of the particle distribution. The un-normalized Sacherer rms emittance is defined by ǫ x, rms = q σ x σ x ′ − σ xx ′ (1)with σ x ≡ p h x i − h x i , σ x ′ ≡ p h x ′ i − h x ′ i , and σ xx ′ ≡ h xx ′ i − h x ih x ′ i . We also use the normalized emit-tances given by ǫ Nx, rms = βγǫ x, rms . (2) B. X-ray Beam Parameters
The total flux of a photon beam, F , is the rate atwhich the photons pass a given location with units ofphotons/sec. The formula specific to a photon beamproduced by inverse Compton scattering will be givenin Sec. III A. The parameter F . represents the flux ina 0.1% bandwidth.The spectral brightness or brilliance of a photon beamis the density of photons in the six-dimensional space con-taining the beam. The general formula for the brillianceof a photon beam into a 0.1% bandwidth is B = F . π σ γ,x σ γ,x ′ σ γ,y σ γ,y ′ (3)where σ γ,x and σ γ,y are the rms transverse sizes of thephoton beam and σ γ,x ′ and σ γ,y ′ are the rms transverseangular sizes of the photon beam. However, by taking ad-vantage of the analogy to undulator radiation, it is possi-ble to approximate the brilliance of the scattered photons using the parameters of the electron beam at the collision.The standard approximation is σ γ,x ′ ≈ p ǫ x /β x + λ/ L ,where ǫ x and β x are parameters of the electron beam, λ is the emitted wavelength, and L is the effective lengthof the source. This result assumes the X-ray beam angu-lar sizes are a combination of the intrinsic beam anglesand radiation diffraction, which is quantified by λ/ L [4].However, the properties of compact sources are such that ǫ x,y > λ/ π , implying that the decrease in brilliance forthe photon source due to λ/ L terms is negligible [6].Taking these approximations into account, Eq. (3) be-comes B = F . π σ γ,x p ǫ x /β x σ γ,y p ǫ y /β y . (4)In previous papers [7–9], we have taken the approxi-mation that the X-ray source size is the size of the elec-tron beam; this approximation is typical in the charac-terization of compact sources [3, 4]. In this approach, σ γ,x = σ x = √ β x ǫ x , so Eq. (4) becomes B ≈ F . π ǫ x ǫ y ≈ γ F . π ǫ Nx, rms ǫ Ny, rms . (5)If instead we take the position that the source size isa convolution of the electron and laser beam sizes, suchthat 1 σ γ = 1 σ laser + 1 σ x σ y . (6)Using this, Eq. (4) becomes B ≈ F . π σ γ p ǫ x ǫ y /β x β y ≈ γ F . π σ γ q ǫ Nx, rms ǫ Ny, rms /β x β y . (7)As the laser spot size becomes increasingly greater thanthe electron beam spot size, the difference betweenEq. (5) and Eq. (7) becomes negligible. However, for thecompact source presented in this paper, the spot sizes areroughly equivalent, making Eq. (7) more appropriate.From either brilliance formula, it becomes clear thatto maximize brilliance requires maximizing the photonflux or electron beam energy or minimizing the electronbeam normalized rms transverse emittances. III. INVERSE COMPTON LIGHT SOURCEA. Inverse Compton Scattering
The process of scattering a photon off an electron atrest is known as Compton scattering. The term inverseCompton scattering (ICS) is used in the situation suchthat the electron loses energy to the incident photons.In the following formulae, Φ is the angle between therelativistic electron and the laser beams, and ∆Θ is theangle between the laser beam and scattered photons. If θ and φ represent spherical polar angles that the scatteredphotons make in the coordinate system such that the electron beam moves along the z axis, then the angle ∆Θis cos ∆Θ = cos Φ cos θ − sin Φ sin θ cos φ . The coordinatesystem is set so the interaction point (IP) of the electronand laser beams occurs in the x − z plane.A general formula expressing the energy of a scatteredphoton in the lab frame, E γ , as a function of the directionof the scattered photon, is E γ (Φ , θ, φ ) = E laser (1 − β cos Φ)1 − β cos θ + E laser (1 − cos Φ cos θ + sin Φ sin θ cos φ ) /E e − (8)where β is the relativistic factor v z /c , E laser is the en-ergy of the typical laser photon, and E e − = γm e c is theenergy of the electron [4]. This formula includes the im-pact of electron recoil. The Thomson formula is a goodapproximation if the electron recoil is negligible, i.e., theenergy of the laser in the beam frame is much less thanthe rest mass of the electron. When this is true, then theformula for the energy of the scattered photon becomes E γ (Φ , θ ) ≈ E laser − β cos Φ1 − β cos θ . (9)It can also be approximated as E γ (Φ , θ ) ≈ E laser γ (1 − β cos Φ)1 + γ θ , (10)where γ is the usual relativistic factor for the electronand γ ≫ π ). The energy of thelaser photon in the beam frame is E ′ laser = γ (1+ β ) E laser .Assuming that the Thomson formula is a good approx-imation, i.e., E ′ laser ≪ mc is true, then the energy ofthe scattered photon is also E ′ laser in the beam frame.Going back into the lab frame, the photons scattered inthe forward (positive z ) direction have the highest en-ergy, which is γ (1 + β ) E laser ≈ γ E laser . The highenergy boundary of emission is called the Compton edge;no radiation is emitted at higher energies. For photonsscattered at the angle θ such that sin θ = 1 /γ (1 /γ ≪ γ E laser , which is also the aver-age energy of the scattered photons. Both the Comptonedge and the number density of scattered photons as afunction of the energy of scattered photons can be seenin Fig. 2.The number of photons produced by scattering an in-cident laser off an electron is proportional to the time-integrated intensity of illumination. Consequently, thetotal photon yield is proportional to the square of thefield strength, as in the case of undulator radiation. Pro-gressing by the analogy with undulator radiation, thefield strength parameter for a plane wave incident laseris defined to be a = eEλ laser πmc , (11) FIG. 2. Number density of scattered photons as a function ofthe energy of scattered photons. Annotated Fig. 2 from [4]. where e is the electron charge, E is the transverse elec-tric field of the laser, λ laser is the laser wavelength, and mc is the rest energy of the electron. This value repre-sents the normalized transverse vector potential for theEM field accelerating the electrons during scattering. ForCompton scattering, a plays a role similar to that of K in the field of undulators. For the case of a ≪
1, thebackscattering is in the linear regime, an assumption thatcontinues as formulae are presented.If we take the assumption that the transverse intensitydistributions of the laser and electron beams are roundGaussian distributions with the rms sizes of σ e and σ laser respectively, then U γ = γ (1 + β ) σ T N e N laser π ( σ e + σ laser ) E laser , (12)where U γ is the total energy of the scattered photonsper collision, N e is the number of electrons in the bunch, N laser the number of photons in the incident laser, and σ T is the Thomson cross section 8 πr e /
3, where r e is theclassical electron radius [4, 10, 11]. The classical electronradius is defined as r e = e / πǫ mc , where e is theelectric charge of the electron, ǫ is the permittivity offree space, m is the mass of the electron, c is the speedof light [12]. From this formula, the total number ofscattered photons N γ is N γ = σ T N e N laser π ( σ e + σ laser ) . (13)Given that the spectral energy density of the scat-tered photons may be analytically computed in the linearThomson backscatter limit, it can further be determinedthat the number of scattered photons within a 0 . N . = 1 . × − N γ .Consequently, the rate of photons (flux) into this band-width is F . = 1 . × − ˙ N γ . For high-frequencyrepetitive sources, ˙ N γ = f N γ , where f is the repetitionrate [4, 10, 11]. B. Compact Sources
There are two main components in an inverse Comptonlight source (ICLS) - a relativistic electron beam and anincident laser. In the last several years, there has beena significant advancement in the technology to producea suitable laser. The details of this progress are largelybeyond the scope of this article, though the status of thecurrent technology will be touched on later. The othercomponent, the focus of this project, is the relativisticelectron beam off which the incident laser scatters.There exist two schemes for accelerating an electronbeam to the desired energy, typically in the range of afew 10s of MeV: a linear accelerator (linac) or a storagering (ring) [3]. A linac is composed of radiofrequency(RF) or superconducting (SC) RF (SRF) cavities thataccelerate the beam to the desired energy [13]. Rings arecircular devices into which a beam of a specific energy isinjected, where the beam may or may not be extractedbefore being used [6].Both of these options have benefits and drawbacks.Existing storage ring projects typically have lower ex-pected fluxes than those of linacs. The expected bright-ness is frequently lower [3], as the smallest achievablenormalized emittances are typically larger for a ring thana linac. Additionally, a full energy linac is often requiredanyway for injection into the ring [3, 13, 14]. However,rings are capable of a high repetition rate, a higher av-erage current than is typical for linacs, and historicallyhave better stability [3, 14].Linac-based ICS X-ray sources have shown promisingresults at lower pulse repetition rates, though these re-sults have yet to be reproduced at higher rates. For elec-tron beams with an energy above 10 MeV, cumbersomeshielding must be included [3, 14]. Current cryogenicequipment for SRF structures, which are utilized in allbut one of the known linac projects (and, indeed, areby some assumed to be necessary for a linac project tosucceed), are more complicated than non-expert usersare comfortable using. Another common feature to mostlinac projects is a superconducting electron gun, a tech-nology with promising results but not yet a mature field [3, 15]. Linac projects are more likely to be capa-ble of shorter bunch lengths, even without compression,smaller normalized emittances, and a greater flexibilityfor phase space manipulations than ring projects [3, 14].Referenced in the literature as the first existing com-pact Compton source is the one built by Lyncean Tech-nologies. An electron beam is produced by a normalconducting linac and injected into a storage ring, whichoccupies a 1 m by 2 m footprint. This machine delivers ∼ ph/s in a 3% energy bandwidth, with the scatteredphoton beam having an rms spot size of ∼ µ m [3, 16].Table I contains some of the current projects with acompact ICLS design. To give some perspective to thesevalues, the best rotating anodes, such as may currently befound in a lab as an X-ray source, have a flux of ∼ × ph/s and a brightness on the order of 10 photons/(sec-mm -mrad -0.1%BW) [14]. On the other hand, an X-raybeam that might typically be found at a large facility hasa flux in the regime of ∼ − ph/s [17] and a bright-ness of ∼ photons/(sec-mm -mrad -0.1%BW) [18].Given these numbers, a robust user program for acompact ICLS machine would require that substantialfluxes of narrow-band X-rays are the desired requirement,rather than the best average brightness. However, the po-tential for such machines, in terms of both performanceand demand, make the prospect of a well-designed com-pact source non-negligible [4]. C. Considerations for This Project and DesignParameters Choices
The main goal of this study was to develop the con-cept for a high-brilliance, high-flux inverse Compton lightsource that would also be relatively affordable and easyto operate by non-experts. High-flux would imply cw op-eration and an SRF linac, and ease of operation wouldsuggest operating with atmospheric helium at 4.2 K orabove. Since the surface resistance of superconductorsincreases quadratically with frequency this would implya low-frequency system. On the other hand, the sizeand cost of the cavities and cryomodules increase as thefrequency is lowered, and a trade-off between the twoconsiderations suggested a frequency range of 300 to 500MHz [7].A number of accelerating structure geometries wereconsidered. The most common and widespread is theTM-type, sometimes referred to as “elliptical”. This ge-ometry is well-understood and has the advantage of hav-ing rotational symmetry. However it was deemed to betoo large in that frequency range. Another type of su-perconducting structure is the spoke geometry which, atthe same frequency, is smaller than the TM-type [26, 27].Several of these cavities have been developed in the fre-quency and velocity range of interest [28]. A 325 MHzsingle-spoke cavity had been successfully developed butwas also deemed to be too large [29]. We finally decidedon a 500 MHz, double-spoke geometry which had also
TABLE I. Comparison of X-ray beam parameters for different ICLS compact designs.Project Type E x (keV) Ph/s Ph/(s-mrad σ γ ( µ m)-mm -0.1%BW)Lyncean at Munich [4, 16, 19] SR 10-20 10 ODU (12 microns) Linac (SC) 12 10 been successfully developed [30]. The spoke geometryhas the disadvantage of introducing quadrupole compo-nents in the electromagnetic fields [31]. As shown later,the contribution of the quadrupole components can bemanaged and its impact on the final performance is min-imal.We would like to emphasize that further advances inthe SRF technology could justify revisiting the geome-try and/or frequency choice but would not invalidate theconclusions of this study.To increase brightness, the normalized rms transverseemittance needs to be minimized, leading to a targetvalue of 0.1 mm-mrad. While this value is considerablysmaller than in other SRF injector guns, as shown in thiswork, a low bunch charge of 10 pC makes this emittanceattainable [7, 15]. To attain a high average flux, consid-ering that the average flux is proportional to both thebunch charge and the repetition rate, a high repetitionrate of 100 MHz was chosen to counterbalance the lowbunch charge. Minimizing the spot size of both electronand laser beams also helped to increase the flux. Thus,the spot size for the electron beam at the IP was set at ∼ µ m, which is small but feasible, though it will requirestate-of-the-art diagnostics at the IP.An electron beam energy of 25 MeV and an incidentscattering laser energy of 1.24 eV were chosen. The cho-sen energies generate X-rays of up to 12 keV. X-raysat 12 keV have a corresponding wavelength of approx-imately one Angstrom, the same as in large third gener-ation synchrotron facilities. For the energy smearing ofthe forward flux to be small relative to the total band-width necessitates that the relative beam energy spreadbe less than 0.03%. At the chosen energy of 25 MeV, thisleads to an rms energy spread requirement of 7.5 keV. Inorder to keep the flux reduction due to the hourglass ef-fect negligible, the compressed bunch length is set to lessthan 1 mm [7].For the best possible X-ray beam, a high quality highpower laser is necessary. The ideal laser would, amongother properties, have a circulating power of 1 MW, com-pared to 100 kW today. One MW is widely regarded asfeasible, but has not yet been achieved in a compact op- tical cavity [3, 4, 7, 32]. The other properties relevantto the optical cavity are less demanding: 1 µ m wave-length (1.24 eV), 5 × ph/bunch, spot size of 3.2 µ mat collision, and peak strength parameter a = 0 .
026 [7].However, a 3.2 µ m laser spot size has an extremely shortRayleigh range (which presents additional challenges), soresults are also presented for a laser spot size of 12 µ m.It is possible to take the properties of the electron beamand incident laser beam and estimate selected parame-ters of the X-ray beam which would be produced from acollision between the two, using formulae presented pre-viously. For a laser spot size of 3.2 µ m, the X-ray beamenergy will be 12 keV with 1 . × photons/bunch. TheX-ray beam flux will be 1 . × ph/s, with an averagebrilliance of 3 × ph/(sec-mm -mrad -0.1%BW). Fora laser spot size of 12 µ m, the X-ray beam energy will be12 keV with 2 . × photons/bunch. The X-ray beamflux will be 2 . × ph/s, with an average brilliance of2 . × ph/(sec-mm -mrad -0.1%BW). These valuesare sufficiently high as to indicate that a compact Comp-ton source which fulfills these parameters is likely to bevery interesting to potential users [3].These specifications are based on and similar to thoseearlier presented in [32]. Desired electron beam parame-ters at the interaction point (IP) are shown in Table II.Optical cavity parameters are shown in Table III, basedon performances that may soon be attainable [4, 32]. Us-ing the values in these tables and the formulae previouslypresented, the resulting X-ray beam can be described bythe quantities in Table IV for the proposed laser spotsizes.One of the benefits of this design is that the layout isentire linear, which can be seen in Fig. 1. This benefitallows for a simpler and more compact design. Whilewe have seen improvement in the transverse emittanceby increasing the length of the bunch off the cathode,a longer bunch length requires a bunch compressor, in-creasing both the size and complexity of the design. Thebunch compressor might be a 3 π or 4 π design, basic ex-amples of which are shown in Fig. 3[9, 33]. Y c oo r d i n a t e ( m ) X coordinate (m) Y c oo r d i n a t e ( m ) X coordinate (m) Beam pathQuadrupoleDipole
FIG. 3. Basic layout examples of 3 π (left) and 4 π (right) bunch compressors. Beam enters at (0, 0).TABLE II. Desired electron beam parameters at interactionpoint.Parameter Quantity UnitsEnergy 25 MeVBunch charge 10 pCRepetition rate 100 MHzAverage current 1 mATransverse rms normalized emittance 0.1 mm-mrad β x,y σ x,y µ mFWHM bunch length 3 (0.9) psec (mm) rms energy spread 7.5 keVTABLE III. Laser parameters at interaction point.Parameter Quantity UnitsWavelength 1 (1.24) µ m (eV)Circulating power 1 MW N γ , Number of photons/bunch 5 × Spot size ( rms ) 3.2, 12 µ mPeak strength parameter, a a = eEλ laser / πmc Repetition rate 100 MHz rms pulse duration 2/3 ps
IV. SRF GUNA. Similar Design Comparison
There exist three types of photoinjectors, or guns,presently: the DC gun, the normal conducting RF gun,and the SRF gun. While the first two types representtechnology that is mature and the result of developmentover many decades, SRF guns are still an emerging tech-nology [14, 15].The concept for an SRF gun was initially publishedin the early 1990s [34], though more consistent publish-ing on the subject did not occur until nearly a decadelater [35–38]. Using the idea of a reentrant cavity for an
TABLE IV. Desired light source parameters.Parameter Laser spot ( µ m) Units3.2 12X-ray energy Up to 12 Up to 12 keVPhotons/bunch 1 . × . × Flux 1 . × . × ph/secAverage brilliance 3 . × . × ph/(s-mm -mrad -0.1%BW)TABLE V. Comparison of various SRF gun designprojects [15, 39].Parameter ODU ICLS NPS WiFEL BNL UnitsFrequency 500 500 200 112 MHzBunch charge 0.01 1 0.2 5 nCTrans. norm. 0.1, 0.13 4 0.9 3 mm-mrad rms emittance SRF gun was first presented in [35], which subsequentlyinspired a number of similar gun designs [15, 39]. Table Vcompares various SRF gun designs with each other andto the parameters ultimately achieved by this study, re-ferred to as ODU ICLS in the table. This table containsthe design parameters for projects at the Naval Post-graduate School (NPS), the University of Wisconsin FEL(WiFEL), and Brookhaven National Lab (BNL).There are two considerations that can be seen fromTable V. The first is that the bunch charge of the ODUICLS gun is smaller than the other designs by an order ofmagnitude or more. The second is that the desired trans-verse normalized rms emittance is also smaller than theother designs by nearly an order of magnitude or more.This reduced bunch charge is what makes the extremelysmall emittance feasible.It is common in RF/SRF gun design to mitigate thegrowth of the transverse emittance of the bunch due tospace charge in order to produce a beam with the small-est emittance. Emittance compensation is the reductionof emittance due to linear space-charge forces [40, 41].One of the most common techniques in emittance com-pensation is the use of a solenoid. By placing a solenoidafter an injector, the goal is to manipulate the transversephase space so that the focusing effect of the solenoidnegates the defocusing effect of the space charge [40–42].This technique is used in the three other SRF gun designslisted in Table V [15].At the beginning of this study, simulations were runthat modeled a bunch exiting the gun which passedthrough a solenoid before entering the linac. This ap-proach to emittance compensation failed in two ways -the transverse normalized rms emittance was not de-creased and the bunch exiting the linac was difficult tomanipulate for compression and final focusing [7]. Con-sequently, in designing the ODU ICLS accelerator a dif-ferent approach was taken, which utilized RF focusingby altering gun geometry to provide focusing, instead ofit being provided by a solenoid as in similar SRF gundesigns [15].RF focusing refers to focusing provided by the RF EMfields of the accelerating structure [43]. One exampleof this is shown in [44], where the RF EM fields of thegun are manipulated by recessing the cathode holder by avarying amount. In Fig. 4, two similar gun geometries areshown, with the only difference between them being therecessed cathode in the bottom right figure. In essence,this alteration to the gun geometry is to produce a radialelectric field which focuses the beam. Ideally, the focus-ing provided will negate the defocusing produced by thespace charge. However, there is a cost to this approach.As the cathode is further recessed, the radial componentof the electric field (and thus the focusing) increases, butthe longitudinal component (which accelerates the beam)decreases [44].By changing the geometry of the nosecone, it is alsopossible to alter the EM fields within the gun. Regardlessof how the radial field is produced, there is still a balanc-ing act that must be found between the accelerating andfocusing fields. Given that increasing the focusing fielddecreases the accelerating one, a simplistic line of thoughtleads one to simply increase the operating gradient un-til the bunch that exits the gun is sufficiently relativisticsuch that space charge is negligible. There are two mainreasons that such an approach is not feasible.First, for any given gun geometry there is a point atwhich increasing the operating gradient is more detri-mental than beneficial to the beam quality. Past thispoint, the strength of the focusing field is actually over-compensating for the effects of space charge on the bunch,increasing the emittance at the exit of the linac. There-fore, in general there exists an operating gradient for agiven geometry which produces the smallest transverseemittance, analogous to choosing the correct lens focallength to focus a beam of light at a particular location.Second, there exists a maximum threshold for surfacefields on an SRF structure for reliable function. As theoperating gradient is directly proportional to the surfacefields, a maximum threshold for the operating gradient
TABLE VI. Bunch distribution off the cathode.Parameter Quantity UnitsLongitudinal distribution PlateauBunch length 4.5 psRise time 1.125 psRadial distribution Uniform rms bunch radius 1 mmInitial transverse momentum 0 mradBunch charge 10 pCInitial kinetic energy 1 keV p z distribution Isotropic exists for any given geometry [9]. B. Initial Bunch Distribution and Drive Laser
The initial bunch distribution off the cathode has theproperties given in Table VI. This bunch is long enoughto make longitudinal space charge effects negligible, whileshort enough to remove the need for a bunch compres-sor, which simplifies the design [9]. In order to produce a4.5 psec flat-top bunch off the cathode, there exist mul-tiple options. One fully realized option is in use in theLCLS injector [45]. This drive laser was manufacturedby
Thales Laser and is a frequency tripled, chirped-pulseamplification system based on a Ti:sapphire laser [45, 46].The specifications called for by the LCLS commissioningrequire a FWHM pulse duration of 6 ps with a repetitionrate of up to 120 Hz. In addition, the laser has an ad-justable pulse duration between 3 and 20 ps [45]. Whilethe pulse duration is in the correct regime this project re-quires, the repetition rate is less than required by nearlytwo orders of magnitude.Another scheme for producing a flat-top bunch off thecathode involves the use of long-period fiber gratings(LPGs). Using this approach, it has been demonstratedexperimentally that Gaussian-like optical pulses can betransformed into flat-top pulses. In the proof of conceptexperiment which confirmed this approach, 600 fs and1.8 ps Gaussian-like pulses were transformed into 1 and3.2 ps flat-top pulses, respectively. The same LPG wasused for both transformations, demonstrating the adapt-ability of such a device [47]. It remains to demonstratethis technology at high average power.
C. Optimization Leading to the Geometry
During the course of the design further gun optimiza-tion was necessary to obtain the desired electron beam atthe IP. To support the optimization it was necessary tocreate a set of parameters to fully define the parametricpiecewise function that describes the gun shape, assum-ing the overall gun shape is retained. A set of formulaewas created that required twelve parameters, shown in R ( c m ) Z (cm) 0 2 4 6 8 10 12 14 0 5 10 15 20 25 30 R ( c m ) Z (cm)
FIG. 4. Two identical gun geometries with (left) and without (right) a recessed cathode to provide RF focusing.TABLE VII. Cavity and RF properties of the gun design. Setto operate at E acc = 10 . E ∗ p B ∗ p B ∗ p /E ∗ p G R/Q ) × G . × Ω Energy content U ∗
44 mJ ∗ At E acc = 1 MV/m Fig. 5.While a cursory examination was made of different pa-rameters, y E is the key parameter to change to producea suitable electron beam at the interaction point. Thisparameter being key is not surprising, given its proximityto both the center of the gun and the cathode holder. Byaltering this parameter over a range of values and evalu-ating the electron beam at the exit of the linac, the gungeometry was chosen. Further optimization of the otherparameters may produce a better design at a later date. D. Final Geometry and Simulation Results
The optimized geometry is shown in Fig. 6, with thephysical and RF properties given in Table VII. We usedIMPACT-T [48] to track 100,000 macroparticles throughthe EM fields simulated by Superfish [49]. The trackingresults at the exit of the gun are shown in Table VIII,with the transverse phase space and beam spot at thegun exit shown in Fig. 7.
TABLE VIII. IMPACT-T tracking results at gun exit.Parameter Quantity Unitskinetic energy 1.51 MeV rms energy spread 0.68 keV σ x,y ǫ N ( x,y ) , rms σ z V. LINACA. SRF Double-Spoke Cavity
Until recently, accelerating electrons near the speedof light has not been attempted with multi-spoke cavi-ties. This is largely because of the well-established andsuccessful performance of TM-type cavities. However,multi-spoke cavities are familiar options for acceleratingions. Previous studies of multi-spoke cavities for β ∼ R (cid:13) [56].Select RF and physical properties are contained in Ta-ble IX. For more information on the optimization of thedouble-spoke cavity design, the reader is directed toward[55]. It is difficult to shorten the linac without requiringgradients which may not be reliably achieved. l fin l gap l rec R cav R cathode R entrance h fin x E y E y E2 R pipe α FIG. 5. Diagram of gun geometry with parameters.FIG. 6. SRF gun geometry.
B. Layout and Simulation Results
One aspect of the double-spoke cavity is the“quadrupole-like” behavior of the cavities - the electronbeam is focused in x and defocused in y , or vice versaby the accelerating mode [13, 31, 57]. This aspect meansthat some adjustment is necessary to provide a roundbeam spot to the bunch compressor or final focusing sec-tion. When arranging the double-spoke cavities, the cen-ter two cavities are rotated 180 ◦ around the y -axis, asseen in Fig. 1. Simulations have demonstrated that thislayout produces the roundest beam at the exit of thelinac.Continuing to simulate the beam past the gun exityields the electron beam properties given in Table X withthe beam spot and phase spaces shown in Fig. 10. Thefinal two cavities are chirped − ◦ off-crest in order toreduce the rms energy spread. At this location, the ex-tremely small transverse normalized rms emittance hasbeen achieved.0 FIG. 7. Beam spot (left), transverse phase space (center), and longitudinal phase space (left) of bunch exiting gun.FIG. 8. The double-spoke SRF cavity, with a portion cutaway to display the interior structure.
C. Emittance Decrease
It has been noted before that the transverse normal-ized rms emittance of the bunch out of the gun is notnecessarily the same out of the linac. In the first itera-tion of the gun design, there was an increase in emittanceafter the bunch exited the gun because it was not yet at asufficient energy to make space charge negligible. In thefinal design, however, the emittance actually decreasesbetween the gun and linac exits. The final iteration hasa greater decrease in emittance and will be examined hereto explain the behavior.The transverse normalized rms emittances and rms spot sizes of the bunch as it passes through the linacare shown in Fig. 11. Both horizontal and vertical emit-tances decrease through the linac, though the rate of de-crease changes with the longitudinal position and whichtransverse component is being considered. The trans-verse rms sizes of the beam grow rapidly immediatelyafter the bunch exits the gun, but the size increase islimited within the linac.Using IMPACT-T, it is possible to see the evolution of -4-2 0 2 4-0.5 -0.25 0 0.25 0.5 E z ( M V / m ) z (m) FIG. 9. The accelerating electric field along the beamline ofthe double-spoke SRF cavity.TABLE IX. Physical (top) and RF (bottom) properties ofdouble-spoke cavity.Parameter Quantity UnitsFrequency of accelerating mode 500 MHzFrequency of nearest mode 507.1 MHzCavity diameter 416.4 mmIris-to-iris length 725 mmCavity length 805 mmReference length [(3/2) β λ ] 900 mmAperture diameter 50 mmEnergy gain ∗ at β
900 kV
R/Q
675 Ω QR † s
174 Ω(
R/Q ) × QR † s × Ω Peak electric surface field E ∗ p B ∗ p B ∗ p /E ∗ p ∗ ∗† ∗ At E acc = 1 MV/m and reference length (3 / β λ , β = 1 † R s = 125 nΩ FIG. 10. Beam spot (upper left), longitudinal phase space (upper right), horizontal phase space (bottom left), and verticalphase space (bottom right) of bunch after exiting the linac. T r a n s v e r s e N o r m a li ze d r m s E m itt a n ce ( mm - m r a d ) z (m)HorizontalVertical 0.28 0.32 0.36 0.4 0.44 1 2 3 4 r m s S po t S i ze ( mm ) z (m)HorizontalVertical FIG. 11. Transverse normalized rms emittances (top) andspot sizes (bottom) of bunch passing through the linac. TABLE X. Properties of electron bunch at linac exit.Parameter Quantity Unitskinetic energy 25. MeV rms energy spread 3.44 keV ǫ Nx, rms ǫ Ny, rms σ x σ y β x
60 m β y
54 m α x -2.3 - α y -3.8 - σ z the bunch after the gun as the beam drifts downstream,without passing through the linac. The transverse nor-malized rms emittance and the spot size of such a driftingbunch are shown as a function of longitudinal position inFig. 12. While the spot size increases as the bunch driftsdownstream, the emittance decreases to a minimum atapproximately z = 0 . z = 0 . ǫ N rms ,r = 0 .
12 mm-mrad at the minimum of2 T r a n s v e r s e N o r m a li ze d r m s E m itt a n ce ( mm - m r a d ) z (m) 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2 0.5 1 1.5 2 r m s S po t S i ze ( mm ) z (m) FIG. 12. Transverse normalized rms radial emittance (top)and transverse spot size (bottom) of bunch drifting after gunexit as a function of longitudinal position. z = 0 . ǫ N rms ,x = 0 .
10 mm-mrad and ǫ N rms ,y = 0 .
13 mm-mrad. So even the averageof the two transverse emittances is less than what canbe attained if the bunch just drifts after the gun. Ifthe bunch charge of the beam exiting the gun is artifi-cially decreased, the distance to the emittance minimumincreases and the emittance minimum decreases. Thiscan be considered analogous to increasing the beam en-ergy without the additional phase space manipulationsof passing the beam through the “quadrupole-like” spokecavities.Increasing the energy of the beam does not mean itis impossible for an emittance minimum to occur withinthe linac; it depends on the bunch exiting the gun. Oneexample of an emittance minimum occurring within thelinac is shown in Fig. 14. The figure shows the transversenormalized rms emittances of the cathode bunch trackedthrough an unoptimized version of the accelerating sec-tion. While the emittances decrease, after the minimumboth increase. At this minimum, ǫ N rms ,x = 0 .
095 mm-mrad and ǫ N rms ,y = 0 .
11 mm-mrad, both of which aresmaller values, respectively, than those of the bunchexiting the linac. With the increase after the mini-mum, the bunch exits with ǫ N rms ,x = 0 .
13 mm-mrad and ǫ N rms ,y = 0 .
13 mm-mrad, so this is not the best possiblesystem for this initial bunch. Consequently, there is somelimit on the rate of emittance decrease for the bunch ex-iting the gun. If the emittance decreases too rapidly,a minimum occurs within the linac, which leads to the
TABLE XI. Select magnet properties of the final focusing sec-tion.Parameter Quantity UnitsMaximum β
132 mQuadrupole length 0.1 mQuadrupole strengths 1.2 - 3.6 T/m rms energy spread 3.4 keVTABLE XII. Select properties of the electron beam parame-ters, both desired and achieved, at the IP.Parameter Desired Achieved Units β x β y ǫ N rms ,x ǫ N rms ,y σ x µ m σ y µ m >
76% longitudinal 3 3 psdistribution rms energy spread 7.5 3.4 keV beam quality suffering.
VI. FINAL FOCUSING
The final focusing section consists of threequadrupoles, with a distance of ∼
29 cm betweenthe third quadrupole and the IP. Tracking was per-formed using elegant and the bunch distribution atthe exit of the linac [58].
Elegant was used in orderto make the optimization easier, but the results werelater compared to tracking using a 3D space chargetracking code and found to have negligible differences.Aberrations of quadrupole displacement were includedin the sensitivity studies. The value of β x and β y areshown as a function of the beam path s in Fig. 15.Certain aspects of the focusing lattice and the propertiesof the bunch at the IP are shown in Tables XI and XII,respectively. The beam spot and phase spaces of thisbeam are shown in Fig. 16. It can be seen that the smallemittance is preserved while focusing the electron beamspot size to a few microns.There is an assumption that the use of spoke cavities inthe linac instead of traditional transverse magnetic (TM)mode cavities, such as multicell elliptical cavities, hasa detrimental effect on the transverse emittance of thebeam. In order to examine this idea, each double-spokecavity in the linac was replaced by a 3-cell elliptical cav-ity; the EM fields were calculated using Superfish. Afterusing IMPACT-T to track the bunch through this ver-sion of the linac, the beam was focused to a small size.The beam spot and phase spaces of the focused bunchis shown in Fig. 17. When compared to the bunch ac-celerated by the double-spoke cavities, seen in Fig. 16,3 FIG. 13. Transverse phase spaces of the bunch as it drifts downstream. T r a n s v e r s e N o r m a li ze d r m s E m itt a n ce ( mm - m r a d ) z (m) HorizontalVertical FIG. 14. Transverse normalized rms emittances of the bunchoff the cathode tracked through an unoptimized version of theaccelerating section as a function of the longitudinal position. the beam is highly comparable. The transverse normal-ized rms emittance of the new bunch is ∼ .
11 mm-mrad,which is the average transverse normalized rms emittanceof the bunch using spoke cavities. Consequently, there isno beam physics reason that elliptical cavities are thebetter option for beam acceleration. ( m ) s (m) β x β y FIG. 15. β x and β y as a function of s in the final focusingsection of the design. The location of the three quadrupolesare positioned along the horizontal axis. VII. SENSITIVITY STUDIES
In order to ascertain the robustness of the design, simu-lations involving deviations from optimal design parame-ters were performed. In these simulations, the maximumthreshold for each perturbation from the design was de-termined to be the point when any electron beam param-4
FIG. 16. Beam spot (top left), longitudinal phase space (top right), horizontal phase space (bottom left), and vertical phasespace (bottom right) of the electron bunch at the IP.TABLE XIII. The amplitude and phase perturbation fromdesign for each SRF structure at which some electron beamparameter changes ∼
20% at the IP.Varied Parameter and Structure ThresholdAmplitude of Gun -2.0%+0.6%Amplitude of All Cavities -1.0%+0.8%Phase of Gun -7.2 ◦ +1.2 ◦ Phase of All Cavities -1.2 ◦ +1.2 ◦ eter value given in Table XI changed by 20%.The phase and amplitude of each SRF structure wasindividually varied while holding all other settings con-stant. The change in phase is given in degrees, while thechange in amplitude is given in percentage of the designvalue. The thresholds are reported in Table XIII. Thelimiting electron beam parameter is the vertical rms sizewhen the amplitude of the cavities is varied. In all othercases, the limiting parameter is the rms energy spread.Systematic perturbations were also evaluated for thecoordinates of both the linac cavities and the magnetsin the final focusing section, separately. In either case,each element was randomly attributed some amount of misalignment in each of the three Cartesian directions.The maximum possible misalignment is the threshold.For the translational (transverse and longitudinal) mis-alignment in the linac cavities, the threshold is 500 µ m,with the limiting electron beam parameter being the rms energy spread. For the translational misalignment of thethree quadrupole magnets, the threshold is 300 µ m, withthe limiting parameter being the vertical size [9]. VIII. INCIDENT LASER
Inverse Compton Light Sources require both an elec-tron beam and an incident laser, the latter of which hasbeen neglected thus far. The design of the appropriatelaser is beyond the scope of this article, but the desiredproperties are provided in Table III. While a laser with acirculating power of 1 MW is called for, such a laser doesnot currently exist. Current consensus of those withinthat field is that such a laser is feasible, but until nowthere has not been a need for it to be developed. Atpresent, high average power lasers currently constructedhave a power of ∼
100 kW. Using a laser with this cir-culating power would decrease the flux and brightness ofthe anticipated X-ray beam by an order of magnitude.5
FIG. 17. Beam spot (top left), longitudinal phase space (top right), horizontal phase space (bottom left), and vertical phasespace (bottom right) of the electron bunch at the IP, using elliptical cavities in the linac.
IX. X-RAY SOURCE
By using the parameter values in Tables XI and III inthe formulae presented in Sec.IX and III A, it is possi-ble to estimate the X-ray beam parameters of the lightsource. These parameters are presented in Table XIV,assuming Gaussian laser and beam spots. However, theelectron distribution at the interaction point is not Gaus-sian, bringing the validity of the results into question.Fortunately, Compton scattering calculations havebeen made recently which utilize the electron beam dis-tribution, not just beam parameters. Using these meth-ods, the calculations of the X-ray light source parame-ters verify that the non-Gaussian distribution does notsignificantly impact the anticipated brilliance [8]. In theprevious paper [8], we used B p = lim θ a → S . π σ e θ a (14)to calculate the pin-hole brilliance of the X-ray beam B p , where S . is the number of photons per secondin a 0.1% bandwidth transmitted through the aperture,which is calculated from the code. However, applying thesame reasoning given in Sec.II B, this formula becomes B p = lim θ a → S . π σ γ θ a . (15) Table XV contains the estimated X-ray beam parameterswhen calculated in this manner, while the full spectrumof the produced X-rays is shown in Fig. 18.The energy and average brilliance in Tables XIV andXV show excellent agreement for the 12 µ m case. Theflux into a 0.1% bandwidth does not, which is expectedbecause this parameter is dependent on the apertureangle of the interaction enclosure. Calculation resultsclearly demonstrate that the flux increases with the aper-ture angle, and the results are being reported for a smallopening. The factor of two increase between the calcu-lated brilliance and the pin-hole brilliance for the 3.2 µ mcase is expected - the calculation code makes the assump-tion that every photon of the scattering laser sees thesame scattering potential, which is most valid when thelaser spot size is much greater than the electron beamspot size. As that assumption is not valid for the 3.2 µ mcase, the code overestimates the anticipated pin-hole bril-liance. X. FURTHER WORK
While we have presented here a preliminary design fora high-brilliance, high-flux inverse Compton light sourcemuch work remains to be done (analytical, numericalsimulation, and experimental) before such a source could6
TABLE XIV. Estimated X-ray performance assuming design electron beam attained at IP, compared to desired parameters.Parameter Laser Spot Size ( µ m) Units3.2 12Desired Achieved Desired AchievedX-ray energy 12 12 12 12 keV N γ . × . × . × . × photons/bunchFlux 1 . × . × . × . × ph/sFlux in 0.1% BW 2 . × . × . × . × ph/(s-0.1%BW)Average Brilliance 3 . × . × . × . × ph/(s-mm -mrad -0.1%BW) d N / d E ( / e V ) E (keV) θ a = 3/20 γθ a = 1/10 γθ a = 1/20 γθ a = 1/40 γ d N / d E ( / e V ) E (keV) θ a = 3/20 γθ a = 1/10 γθ a = 1/20 γθ a = 1/40 γ FIG. 18. Number spectra for different apertures generated using 4,000 particles for a 3.2 µ m laser spot size (left) and a 12 µ mlaser spot size (right). Grey box indicates 0.1% bandwidth. Suggested apertures for brilliance calculation only.TABLE XV. X-ray performance of the designs attained bynumerical simulation with an aperture of 1/40 γ . Suggestedaperture for brilliance calculation only.Parameter Laser Spot ( µ m) Units3.2 12X-ray energy 12.3 12.3 keV N . S . . × . × ph/(s-0.1%BW) B p . × . × ph/(s-mm -mrad -0.1%BW) be built. In particular, it may be that the choices of fre-quency and geometry are not optimal and may be revisedafter further study or advances is the SRF technology. Inthis section we outline the major items that would requirefurther R&D activities. A. SRF Gun
A number of simulation studies need to be conductedbefore the SRF gun is built and commissioned - primarilymultipacting and mechanical. If geometry alterations aredeemed necessary, another optimization based on simula-tion beam dynamics results may be necessary, includingthese studies. Once these studies are completed and the geometry isfinalized, the gun will need to be built and commissioned.At this point, experiments can be performed to demon-strate that the transverse normalized rms emittance be-havior as the beam drifts after the gun is as expected,supporting the idea that appropriate gun geometry canprovide all necessary emittance compensation.Integration of a photo-cathode and an SRF gunpresents a technical challenge which is under investiga-tion in a number of institutions.
B. Beam Physics
Further sensitivity studies are called for, especially ifthe gun geometry needs to be altered to avoid significantobstacles in multipacting, mechanical stability, or ther-mal breakdown. Additionally, simulations examining po-tential wakefields are desired. All these simulations willneed to include potential deviations from an ideal design(physical misalignment, errors of amplitude and phase,etc).7
C. SRF Linac
The design presented here is based on a particularchoice of frequency and geometry in order to achieve abalance between size (capital cost) and operating cost.This choice may not be optimal and, in particular, theoperating cost would still be higher than desired basedon the surface resistance assumed in Table IX. Howeverrecent progress in SRF R&D shows potential for a sub-stantial reduction of power dissipation in SRF cavities.Nitrogen doping [59] and infusion [60] during heat treat-ment have shown large reduction of those losses at 2 Kand higher frequency. Achieving similar results at 4.2 Kand 500 MHz would validate our choices. On a longerterm, Nb Sn [61] could offer dramatic reduction in cryo-genic losses at 4.2 K and would even allow operation athigher frequency; such an advance would lead us to re-visit our choices as multi-cell TM-type cavities operatingaround 650 MHz would be attractive. They may evenbe able to operate without a refrigerator, using insteadcryocoolers [62] if vibrations that such systems often gen-erated can be managed.Eventually, several prototype cavities will need to bebuilt and tested, and all the processes needed to achieveperformance (chemistry, heat treatment, cleaning, etc)will have to be developed and demonstrated.Performance of this light source is contingent uponachieving and maintaining a very small emittance. Thiswill put challenging constraints on the design of the cry-omodule and the Low Level RF Control system.
D. Incident Laser
As previously mentioned, a laser with the desired prop-erties of either spot size does not currently exist. Whilecurrent technology might suffice in providing an X-raybeam at least as good as any other compact ICLS, thisdesign has the capacity to surpass that threshold. Conse-quently, such a laser must be designed and commissioned,before the project presented here is commissioned.The benefits of developing a better incident laser donot stop with this project, however. Other compact ICLSprojects, proposed or existing, can see an improvementin the quality of the X-ray beam they deliver. This ap-proach may be one method which will see such develop-ment funded. Additionally, other applications for a morepowerful laser do exist.
E. Overall
A complete design should be produced, including allnecessary components - klystrons, cathode drive laser, re-frigeration support, beam dump, etc. The commissioningprocess itself is likely to be divided into two main stages- the electron beam and the X-ray beam. Initially, theelectron beam will need to be produced at the intended interaction point, with the intended properties. After-wards, the incident scattering laser will be installed, withthe appropriate beam transport to allow the produced X-ray beam to travel to the users and the electron beam totravel to the beam dump.
XI. SUMMARY
The Compact ICLS design presented here would im-prove on all other compact sources to date, producingan X-ray beam of quality which is closer than ever tobeing comparable to beams produced at large-scale facil-ities. This is made possible by using cw superconductingRF to accelerate the beam before it is focused to the in-teraction point. At the interaction point, the electronbeam has a small spot size and small transverse normal-ized rms emittance, which correspondingly result in anX-ray beam with high flux and brilliance. The ultra-lowemittance is made feasible by a low bunch charge, witha high repetition rate so the X-ray flux is not adverselyaffected.This design achieves an electron bunch which generatesan X-ray beam unmatched in quality by other CompactICLS designs. These desired electron beam parametersare achieved by utilizing a number of different techniques.The most effective technique was the emittance compen-sation by RF focusing. By altering the geometry of thegun to provide the correct RF focusing for a given bunch,it is possible to produce bunches with low normalizedtransverse rms emittances. Taken together with the lowbunch charge, the achieved transverse emittances are suf-ficiently small. Choosing the correct bunch length off thecathode is necessary, in order to produce a bunch exit-ing the linac which does not need compression, but isstill long enough that the transverse space charge effectscan be compensated for by the RF focusing provided bythe gun geometry. Another beneficial technique is takingadvantage of the quadrupole-like behavior of the double-spoke cavities which comprise the linac in order to pro-duce a fairly round beam at the exit of the linac. Anapproximately round beam at the exit to the linac allowsfor the bunch to be easily focused down to a small spotsize on the order of a few microns.While the incident laser remains an incomplete compo-nent, its design should not be an obstacle. Further workoutside of this aspect includes further optimization toimprove on the current X-ray parameters or altering thedesign for different functions - by increasing the X-rayenergy, for example.
ACKNOWLEDGMENTS
The authors would like to thank Rocio Olave, KarimHern´andez-Chah´ın, Subashini De Silva, ChristopherHopper, Randika Gamage, and Todd Satogata for ear-lier contributions to this project, and HyeKyoung Park8for a discussion of cavity surface treatments. This mate-rial is based on work supported by the U. S. Departmentof Energy, Office of Science, Office of Nuclear Physicsand Office of Basic Energy Science; and by the NationalScience Foundation. K.E.D and J.R.D. were supportedat ODU by DOE Office of Nuclear Physics award No.de-sc00004094. B.T. was supported at ODU by NSF Award 1535641. G.A.K was supported at Jefferson Labby contract DE-AC05-06OR23177; additional supportwas provided by DOE Office of Nuclear Physics AwardNo. DE-SC004094 and Basic Energy Sciences Award No.JLAB-BES11-01. This research used resources of the Na-tional Energy Research Scientific Center, which is sup-ported by the Office of Science of the U.S. DOE underContract No. DE-AC02-05CH11231. [1] W. C. R¨ontgen, Nature (London) , 274 (1896).[2] M.-E. Couprie and J.-M. Filhol,C. R. Physique , 487 (2008).[3] M. Jacquet, Nucl. Instrum. Methods B , 1 (2014),11th European Conference on Accelerators in AppliedResearch and Technology.[4] G. A. Krafft and G. Priebe,Reviews of Accelerator Science and Technology , 147 (2010).[5] K. Dietrick, J. R. Delayen, and G. K. Krafft, in Proc. of the 8th International Particle Accelerator Conference, IPAC 2017, Geneva, Switzerland (JACoW, Geneva, Switzerland, 2017) p. 932.[6] A. Chao, K. Mess, M. Tigner, and F. Zimmermann,
Handbook of Accelerator Physics and Engineering (WorldScientific, 2013).[7] K. Deitrick, J. R. Delayen, B. R. P. Gamage, K. Hernan-dez, C. Hopper, G. A. Krafft, R. Olave, and T. Satogata,(2013).[8] G. A. Krafft, E. Johnson, K. Deitrick, B. Terzi´c, R. Kel-mar, T. Hodges, W. Melnitchouk, and J. R. Delayen,Phys. Rev. Accel. Beams , 121302 (2016).[9] K. Deitrick, Inverse Compton Light Source: A CompactDesign Proposal , Ph.D. thesis, Old Dominion University(2016).[10] J. Clarke,
The Science and Technology of Undulatorsand Wigglers (Oxford Series on Synchrotron Radiation,2004).[11] P. Duke,
Synchrotron Radiation: Production and Prop-erties (Oxford Science, 2000).[12] J. Jackson,
Classical Electrodynamics (John Wiley andSons, 1999).[13] H. Wiedemann,
Particle Accelerator Physics (Springer,2007).[14] W. S. Graves, J. Bessuille, P. Brown, S. Carbajo,V. Dolgashev, K.-H. Hong, E. Ihloff, B. Khaykovich,H. Lin, K. Murari, E. A. Nanni, G. Resta, S. Tantawi,L. E. Zapata, F. X. K¨artner, and D. E. Moncton,Phys. Rev. ST Accel. Beams , 120701 (2014).[15] A. Arnold and J. Teichert,Phys. Rev. ST Accel. Beams , 024801 (2011).[16] “Lyncean technologies, inc. - synchrotron light and the lyncean compact light source,” .[17] “Beamlines directory — advanced photon source,” .[18] S. Benson, M. Borland, D. Douglas, D. Dowell,C. Hernandez-Garcia, D. Kayran, G. Krafft, R. Legg,E. Moog, T. Obina, R. Rimmer, and V. Yakimenko,Nucl. Instrum. Methods A , 1 (2011).[19] E. Eggl, M. Dierolf, K. Achterhold, C. Jud,B. G¨unther, E. Braig, B. Gleich, and F. Pfeiffer,Journal of Synchrotron Radiation , 1137 (2016).[20] P. Yu and W. Huang,Nucl. Instrum. Methods A , 1 (2008).[21] E. G. Bessonov, M. V. Gorbunkov, P. V. Kostryukov, Y. Y. Maslova, V. G. Tunkin, A. A. Postnov, A. A.Mikhailichenko, V. I. Shvedunov, B. S. Ishkhanov, andA. V. Vinogradov, J. Instrum. , P07017 (2009).[22] A. Variola and T. Collaboration, in Proc. of the 2nd International Particle Accelerator Conference, IPAC’11, San Sebastian, Spain (JACoW, Geneva, Switzerland, 2011) p. 1903.[23] J. Urakawa, Nucl. Instrum. Methods A , S47 (2011),the International Workshop on Ultra-short Electron &Photon Beams: Techniques and Applications.[24] R. Hajima, N. Kikuzawa, N. Nishimori,T. Hayakawa, T. Shizuma, K. Kawase, M. Kando,E. Minehara, H. Toyokawa, and H. Ohgaki,Nucl. Instrum. Methods A , S57 (2009).[25] E. Bulyak, P. Gladkikh, A. Zelinsky, I. Kar-naukhov, S. Kononenko, V. Lapshin, A. Myt-sykov, Y. Telegin, A. Khodyachikh, A. Shcherbakov,V. Molodkin, V. Nemoshkalenko, and A. Shpak,Nucl. Instrum. Methods A , 241 (2002).[26] J. R. Delayen, in
Proc. of the 2010 Linear Accelerator Conference, LINAC10, Tsukuba, Japan (JACoW, Geneva, Switzerland, 2010) p. 377.[27] J. R. Delayen, in
Proc. of the 2012 Linear Accelerator Conference, LINAC12, Tel Aviv, Israe (JACoW, Geneva, Switzerland, 2012) p. 758.[28] C. S. Hopper, H. Park, and J. R. Delayen, in
Proc. of the 2014 Linear Accelerator Conference, LINAC14, Geneva, Switzerland (JACoW, Geneva, Switzerland, 2014) p. 677.[29] C. S. Hopper, H. Park, and J. R. Delayen, in
Proc. of the 2015 International Conference on RF Superconductivity, SRF 2015, Whistler, Canada (JACoW, Geneva, Switzerland, 2015) p. 744.[30] H. Park, C. S. Hopper, and J. R. Delayen, in
Proc. of the 2014 Linear Accelerator Conference, LINAC14, Geneva, Switzerland (JACoW, Geneva, Switzerland, 2014) p. 385.[31] C. S. Hopper, K. Deitrick, and J. R. Delayen, in
Proc. of the 2013 North American Particle Accelerator Conference, NAPAC2013, Pasadena, CA, USA (JACoW, Geneva, Switzerland, 2013) p. 868.[32] W. Graves, W. Brown, F. Kaertner, and D. Moncton,Nucl. Instrum. Methods A , S103 (2009).[33] B. R. P. Gamage and T. Satogata, in
Proc. of the 2013 North American Particle Accelerator Conferenc,e NAPAC2013, Pasadena, CA, USA (JACoW, Geneva, Switzerland, 2013) p. 1346.[34] A. Michalke,
Photocathodes Inside SuperconductingCavities-Studies on the Feasibility of a Superconduct-ing Photoemission Souce of High Brightness , Ph.D. the-sis, Bergische Universit¨at Gesamthochschule Wuppertal(1993).[35] A. Michalke, H. Piel, C. Sinclair, P. Michelato,C. Pagani, L. Serafini, and M. Peiniger, in
Proc. of the 5th Workshop on RF Superconductivity, SRF1991, Hamburg, Germany (JACoW, Geneva, Switzerland, 1991) p. 734.[36] A. Michalke, H. Piel, C. Sinclair, and P. Michelato, in
Proc. of the 3rd European Particle Accelerator Conference, IPAC1992, Berlin, Germany (JACoW, Geneva, Switzerland, 1992) p. 1014. [37] C. Pagani, P. Michelato, and L. Serafini,Nucl Instrum. Methods A , 17 (1994).[38] D. Janssen, H. Bttig, P. Evtushenko, M. Freitag,F. Gabriel, B. Hartmann, U. Lehnert, P. Michel,K. Mller, T. Quast, B. Reppe, A. Schamlott,C. Schneider, R. Schurig, J. Teichert, S. Konstantinov,S. Kruchkov, A. Kudryavtsev, O. Myskin, V. Petrov,A. Tribendis, V. Volkov, W. Sandner, I. Will, A. Math-eisen, W. Moeller, M. Pekeler, P. Stein, and C. Haber-stroh, Nucl. Instrum. Methods A , 314 (2003), pro-ceedings of the 24th International Free Electron LaserConference and the 9th Users Workshop.[39] J. R. Harris, K. L. Ferguson, J. W. Lewellen, S. P. Niles,B. Rusnak, R. L. Swent, W. B. Colson, T. I. Smith,C. H. Boulware, T. L. Grimm, P. R. Cunningham, M. S.Curtin, D. C. Miccolis, D. J. Sox, and W. S. Graves,Phys. Rev. ST Accel. Beams , 053501 (2011).[40] B. Carlsten, Nucl. Instrum. Methods A , 313 (1989).[41] L. Serafini and J. B. Rosenzweig,Phys. Rev. E , 7565 (1997).[42] V. Kumar, Am. J. Phys. , 737 (2009),http://dx.doi.org/10.1119/1.3129242.[43] T. Wangler, RF Linear Accelerators (John Wiley andSons, Inc., 2008).[44] D. Janssen and V. Volkov,Nucl. Instrum. Methods A , 34 (2000).[45] R. Akre, D. Dowell, P. Emma, J. Frisch, S. Gile-vich, G. Hays, P. Hering, R. Iverson, C. Limborg-Deprey, H. Loos, A. Miahnahri, J. Schmerge,J. Turner, J. Welch, W. White, and J. Wu,Phys. Rev. ST Accel. Beams , 030703 (2008).[46] D. H. Dowell, J. Castro, P. Emma, J. Frisch,S. Gilevich, G. Hays, P. Hering, C. Limborg-Deprey, H.Loos, A. Miahnahri, and W. White, in Proc. of Particle Accelerator Conference, PAC07, Albuquerque, NM, USA ,Particle Accelerator Conference No. 22 (IEEE, Piscat-away, NJ, USA, 2007) pp. 1317–1319.[47] Y. Park, M. Kulishov, R. Slav´ık, and J. Aza˜na,Opt. Express , 12670 (2006).[48] J. Qiang, S. Lidia, R. D. Ryne, and C. Limborg-Deprey,Phys. Rev. ST Accel. Beams , 044204 (2006).[49] “Poisson superfish,” (2012). [50] C. S. Hopper, R. G. Olave, and J. R. Delayen, in Proc. of the 4th International Particle Accelerator Conference, IPAC 2013, Shanghai,China (JACoW, Geneva, Switzerland, 2013) p. 2480.[51] C. S. Hopper, R. G. Olave, and J. R. Delayen, in
Proc. of the 3rd International Particle Accelerator Conference, IPAC 2012, New Orleans, LA, USA (JACoW, Geneva, Switzerland, 2012) p. 2456.[52] C. S. Hopper and J. R. Delayen, in
Proc. of the 15th International Conference of RF Superconductivity, SRF2011, Chicago, IL, USA (JACoW, Geneva, Switzerland, 2011) p. 154.[53] J. R. Delayen, S. U. De Silva, and C. S. Hopper, in
Proc. of the 2011 Particle Accelerator Conference, PAC 2011, New York, NY, USA (JACoW, Geneva, Switzerland, 2011) p. 1024.[54] C. S. Hopper,