Document [original]

Struct. Dyn. 4, 014301 (2017); https://doi.org/10.1063/1.4976004 4, 014301

Multi-color imaging of magnetic Co/Pt

heterostructures

Cite as: Struct. Dyn. 4, 014301 (2017); https://doi.org/10.1063/1.4976004

Submitted: 20 December 2016 . Accepted: 30 January 2017 . Published Online: 16 February 2017

Felix Willems, Clemens von Korff Schmising, David Weder, Christian M. Günther, Michael Schneider,

Bastian Pfau, Sven Meise, Erik Guehrs, Jan Geilhufe, Alaa El Din Merhe, Emmanuelle Jal, Boris Vodungbo,

Jan Lüning, Benoit Mahieu, Flavio Capotondi, Emanuele Pedersoli, David Gauthier, Michele Manfredda,

and Stefan Eisebitt

ARTICLES YOU MAY BE INTERESTED IN

Perspective: Ultrafast magnetism and THz spintronics

Journal of Applied Physics 120, 140901 (2016); https://doi.org/10.1063/1.4958846

Generating circularly polarized radiation in the extreme ultraviolet spectral range at the free-

electron laser FLASH

Review of Scientific Instruments 88, 053903 (2017); https://doi.org/10.1063/1.4983056

Probing ultrafast changes of spin and charge density profiles with resonant XUV magnetic

reflectivity at the free-electron laser FERMI

Structural Dynamics 4, 055101 (2017); https://doi.org/10.1063/1.4990650

Multi-color imaging of magnetic Co/Pt heterostructures

Felix Willems,

Clemens von Korff Schmising,

1,a)

David Weder,

Christian M. G€

unther,

Michael Schneider,

Bastian Pfau,

Sven Meise,

Erik Guehrs,

Jan Geilhufe,

Alaa El Din Merhe,

3,4

Emmanuelle Jal,

3,4

Boris Vodungbo,

3,4

Jan L€

uning,

3,4

Benoit Mahieu,

Flavio Capotondi,

Emanuele Pedersoli,

David Gauthier,

Michele Manfredda,

and

Stefan Eisebitt

1,2

Max-Born-Institute Berlin, 12489 Berlin, Germany

Institut f€

ur Optik und Atomare Physik, Technische Universit€

at Berlin, 10623 Berlin,

Germany

Sorbonne Universit

es, UPMC Universit

e Paris 06, UMR 7614, LCPMR, 75005 Paris,

France

CNRS, UMR 7614, LCPMR, 75005 Paris, France

Laboratoire d’Optique Appliquee, ENSTA ParisTech, CNRS, Ecole Polytechnique,

Universit

e Paris-Saclay, 828 boulevard des Mar

echaux, 91762 Palaiseau Cedex, France

Elettra-Sincrotrone Trieste, 34149 Basovizza, Trieste, Italy

(Received 20 December 2016; accepted 30 January 2017; published online 16 February

2017)

We present an element specific and spatially resolved view of magnetic domains

in Co/Pt heterostructures in the extreme ultraviolet spectral range. Resonant

small-angle scattering and coherent imaging with Fourier-transform holography

reveal nanoscale magnetic domain networks via magnetic dichroism of Co at the

2,3

edges as well as via strong dichroic signals at the O

2,3

and N

6,7

edges of Pt.

We demonstrate for the first time simultaneous, two-color coherent imaging at a

free-electron laser facility paving the way for a direct real space access to

ultrafast magnetization dynamics in complex multicomponent material systems.

C2017 Author(s). All article content, except where otherwise noted, is licensed

under a Creative Commons Attribution (CC BY) license (http://

creativecommons.org/licenses/by/4.0/).[http://dx.doi.org/10.1063/1.4976004]

I. INTRODUCTION

Magnetic systems with Co/Pt interfaces exhibit a wealth of intriguing phenomena based on

strong spin orbit interaction. Some recent examples include the control of domain wall motion

arising from Dzyaloshinskii-Moriya interaction and spin Hall currents,

room temperature

dynamics of skyrmions in a magnetic racetrack geometry

and ultrafast, all-optical control of

electric currents in ferromagnetic heterostructures

and present promising new opportunities for

spintronic devices based on Co/Pt sample systems. Furthermore, femtosecond optical excitation

of bilayers of ferromagnetic and nonferromagnetic layers have been shown to induce an effi-

cient spin-to-charge conversion via the inverse Hall effect

and have led to efficient ultrabroad-

band emitters of terahertz radiation.

Finally, all-optical helicity-dependent switching in the

technologically important class of Co/Pt multilayers and FePt granular thin films

has triggered

an intense debate discussing the responsible microscopic processes.

7–9

In particular, the hypoth-

esis that the optically induced switching is triggered by an initial stochastic nucleation process

in form of mesoscopic magnetic domain structures

10,11

calls for novel experimental techniques

that give a direct and simultaneous access to the element specific magnetization with nanometer

spatial and femtosecond temporal resolution.

[email protected]

2329-7778/2017/4(1)/014301/12 V

CAuthor(s) 20174, 014301-1

STRUCTURAL DYNAMICS 4, 014301 (2017)

Novel light sources like high harmonic generation (HHG) and free electron lasers (FELs)

generate a brilliant radiation covering the spectral range from the extreme ultraviolet (XUV) to

the soft X-ray region with unique properties regarding its ultrashort temporal pulse structure for

femtosecond time resolution, its tunable photon energies for element-selective spectroscopy and

its high degree of spatial coherence for nanoscale imaging techniques. Additionally, laser gener-

ated high harmonic spectra and novel two-color schemes at free electron laser facilities

12,13

allow simultaneous probing of different elements of complex materials.

In this contribution, we present a magnetic small-angle scattering (SAXS) and Fourier

transform holography (FTH) experiments of Co/Pt heterostructures in the XUV spectral range.

Strong magnetic scattering cross sections exist at both the Co M

2,3

edges as well at the O

2,3

and N

6,7

edges of Pt, leading to corresponding bright 1st order diffraction in SAXS and high-

contrast and high-resolution real space images in FTH. These results allow us to design and

carry out the first two-color coherent imaging experiment at the free-electron laser facility

FERMI, where a single hologram encodes the real space information of the magnetic domain

network stemming from Co and Pt.

II. MAGNETIC RESONANT SMALL-ANGLE SCATTERING

The performed magnetic resonant small-angle scattering experiment serves to determine

the amplitude of the magnetic scattering cross section as a function of energy as well to deter-

mine the average length scale of the magnetic nanostructure. This allows a fast benchmark of

the sample system and identifies the optimal energy range for the increasingly complex coher-

ent single- and two-color imaging experiments.

The experimental setup of the SAXS experiment is schematically shown in Figure 1(a).The

energy dependent small-angle scattering experiment was performed at the synchrotron facility

BESSY II at the undulator beamline UE112-PGM.

The number of photons in the energy range

between 35 eV and 80 eV is on the order of 10

ph/s; a monochromator yields a maximal energy

resolution of E/DE>20.000. The Co/Pt multilayer (Fig. 1(d)), with a composition of Al(10)/

FIG. 1. Schematic of the experimental setup for (a) resonant small-angle X-ray scattering (SAXS) and (e) magnetic Fourier

transform holography (FTH). In the SAXS experiment, the magnetic domains are aligned in stripes (b) and lead to bright

first order scattering peaks centered at a momentum transfer 6q(c). In the FTH experiment, the magnetic domains exhibit

a labyrinth network (f) leading to an isotropic magnetic small angle scattering pattern in the difference image between left

and right circular polarization (I

rþ

I

r

) (g). The corresponding sample compositions and geometries are shown on the

right hand side ((d) and (h)).

014301-2 Willems et al. Struct. Dyn. 4, 014301 (2017)

Pt(2)/[Co(0.6/Pt(0.8))]

/Al(3) nm, and out-of-plane anisotropy was deposited on a Si

mem-

brane (50 lm50 lm30 nm) by magnetron sputtering. Prior to the small-angle scattering

experiment, the magnetic domains were aligned to form a stripe geometry by applying an oscillat-

ing, successively decreasing in-plane external magnetic field (Fig. 1(b)).

The alternating magne-

tized domains have an opposite dichroic index of refraction at core-hole transitions such that the

sample acts as a magnetic diffraction grating.

16,17

The advantage of a stripe pattern for magnetic

SAXS experiments is manifold: first of all it leads to well defined diffraction spots in comparison

to a spread out ring diffraction pattern predicted for an isotropic labyrinth domain network. This

leads to an improved signal to noise ratio, without influencing the energy dependence of the mag-

netic scattering intensity. Furthermore, the two unused quadrants of the charged-coupled device

(CCD) detector allow to simultaneously collect scattering from additional grating structures inte-

grated into the sample substrate for XUV beam intensity monitoring.

Finally, we avoid an over-

lap of the scattering pattern with the beam stop. The sample was placed close to the focus of the

XUV beam, and the scattering pattern was recorded with a back-illuminated charged-coupled

device (CCD) placed D

SAXS

¼74 mm downstream of the sample, sufficiently close to detect the

first order diffraction peaks for the smallest energy at 35 eV. The direct beam and charge scatter-

ing of the membrane edges were blocked by a beam stop. The polarization of the XUV radiation

was set to negative circular helicity. We set the integration time to 1 s to avoid saturation of the

CCD detector and accumulated 4 images for each photon energy between 35 eV to 80 eV in

0.5 eV steps.

The scattering pattern for a photon energy of E

¼60 eV, resonant at the Co M

2,3

edge, is

shown in Figure 1(c) and exhibits two bright spots indicating that the magnetic domains are

indeed in a well aligned domain state. In Figure 2(a), we plot the azimuthally integrated scatter-

ing intensity for XUV photon energies of 52 eV, 60 eV, and 72 eV as a function of the momen-

tum transfer q, which is calculated as

q¼4p

ksin H

ðÞ

¼4pEPh

hc sin 1

2atan dpix r

DSAXS



:(1)

FIG. 2. (a) Azimuthally integrated small-angle XUV scattering of an aligned magnetic domain network resonant at the Co

2,3

edge at E

¼60 eV, Pt O

and N

edge at E

¼52 and 72 eV, respectively, as a function of the scattering vector q.

The solid lines are non-linear least square fits to a pseudo-Voigt function centered at q

¼41 lm

1

with a full width at

half maximum of Dq

¼6.6 lm

1

. (b) Energy spectrum of the total number of scattered photons determined by calculat-

ing the area of the corresponding pseudo-Voigt functions. We identify the pronounced magnetic dichroism at the Co M

2,3

edge as well at the Pt O

2,3

and N

6,7

transitions. Note that the scattering intensity at the Pt N

edges is spectrally very narrow

and that its peak value is significantly larger than at the Co M

2,3

edge. The solid line is a guide to the eye.

014301-3 Willems et al. Struct. Dyn. 4, 014301 (2017)

(k) is the XUV photon energy (wavelength), hthe scattering angle, d

pix

the side length of a

CCD pixel (13.5 lm), and rthe radius in pixels along which each azimuthal integration is per-

formed. The qvalues are corrected for the small deviation of the planar CCD detector from a

sphere to the absolute momentum transfer in reciprocal space. The profiles are well described

by a pseudo-Voigt profile, a non-linear least square fit (solid lines in Figure 2(a)) determines a

constant center at q

¼40.9 lm

1

with a full width at half maximum of Dq

¼6.6 lm

1

These values are constant over the entire measured energy range with a standard deviation of

0.1 lm

1

. This corresponds to an average magnetic domain periodicity of d

¼2p/q

¼(154 61) nm. The area of the pseudo-Voigt function is a measure for the total scattering

intensity and is shown in Figure 2(b) as a function of the photon energy. We identify 5 distinct

intensity maxima which we can assign to the following magnetically dichroic resonances: Co

2,3

(3p

1/2,3/2

!3d) centered around 60 eV, Pt O

2,3

(5p

1/2,3/2

!3d) at 66 eV and 52 eV and Pt

6,7

(4f

5/2,7/2

!5d) at 75 eV and 72 eV, respectively. These values are in qualitative agreement

with previously measured magnetic circular dichroism (XMCD) spectra.

19–21

However, one

needs to keep in mind that the magnetic domains act as both a magnetic phase and an absorp-

tion grating. Note that the signal at the Pt N

edge has a very narrow spectral width below our

energy step width of 0.5 eV and significantly exceeds the scattering intensity of the Co M

2,3

edge. Since the resonantly scattered intensity is proportional to the square of the magnetic

structure factor,

16,22

this small-angle XUV scattering experiment acts as a very sensitive probe

for the magnetization with element specificity and access to nanometer spatial resolution. In the

present case, we measure the lateral spatial profiles of the magnetization in the Co layers as

well as the induced magnetization in Pt; the energy independent momentum transfer q

clearly

indicates a laterally homogeneous magnetization of the entire multilayer. Spatial separation of

the scattering peaks in a two- or multi-color experiment for simultaneous element specificity at

Co and Pt can be achieved by adapting the sample detector distance D

SAXS

. Furthermore, it is

noteworthy that the Pt N

resonance is below the Al L-edge, which allows the use of Al metal-

lic filters against visible and infrared radiation for time resolved, optical pump-XUV probe stud-

ies using free electron laser or high harmonic radiation.

17,23

Because the induced magnetization

of Pt is known to be confined to a few atomic layers at the Co/Pt boundary,

19,24

envisioned

time resolved experiments will hence not only track the lateral spatial magnetization profiles

after ultrafast laser excitation, but will also give a detailed view on the physics of interface

magnetism.

III. FOURIER TRANSFORM HOLOGRAPHY

Fourier-transform holography encodes the real-space information of the magnetic nano-

structure by interference of the magnetic small-angle scattering stemming from the object with

a known reference wave. Because of this direct connection between SAXS and FTH (cf.

Figures 1(c) and 1(g)), we take the measured energy spectrum of the SAXS intensity (Figure

2(b)) to infer the optimal energy range for which we can image the element specific magnetiza-

tion of Co and Pt with a maximal signal to noise ratio.

The experimental geometry for the coherent imaging experiments via Fourier transform

holography is shown in Figure 1(e). The measurement was also performed at the undulator

beamline UE112-PGM

of the synchrotron facility BESSY II (HZB). The sample was placed

approximately 100 mm behind the focus, where we confirmed a high degree of transversal coher-

ence via a Young double slit interference experiment.

We determined a coherence jl

j>90%

for a slit separation of 18 lm, significantly exceeding the object-reference distance of j~

rj¼5lm.

The holographic mask was manufactured in a standard transmission configuration:

26–28

A sili-

con nitride membrane (thickness 30 nm) supported by a silicon frame acted as a substrate. After

evaporation of an XUV opaque gold film (thickness 250 nm, maximal transmission <10

8

for

energies between 57 eV and 76 eV), the field of view is defined by drilling a circular object

hole with a diameter of dr¼2lm via ion-beam lithography. Subsequently, the magnetic multi-

layer film Al(10)/Pt(2)/[Co(0.6)/Pt(0.8)]

/Al(3) nm was deposited via magnetron sputtering,

and finally 5 reference holes with 60 nm and 80 nm diameter were added (Figure 1(h)). The

014301-4 Willems et al. Struct. Dyn. 4, 014301 (2017)

coherent scattering from the object and the reference holes interfere with the CCD camera

placed D

FTH

¼53 mm behind the sample and form the intensity hologram. Due to the limited

dynamic range of the camera, we block the direct beam by a circular beam stop. We recorded

five holograms for positive, r

, and negative, r

–

, helicity for XUV photon energies ranging

from 57 eV to 65 eV and from 71 eV to 75.5 eV in 0.5 eV steps. The integration time for a sin-

gle hologram varied from 20 to 25 s making full use of the dynamic range of the detector. The

difference hologram (I

rþ

I

rþ

) contains only magnetic contributions, an example for

¼60 eV is shown in Figure 1(g). The hologram exhibits pronounced magnetic speckles in

the small-angle scattering signal due to the coherent illumination of the masked sample area. In

addition, strong intensity fringes from the object–reference interference appear with a period of

approximately 16 pixels extending all the way to the edge of the detector. The selected energy

range around the Co M and Pt N edges for the imaging experiment was chosen according to

the SAXS measurement (Figure 2(b)) combining the maximal scattering strength as well as the

best element specificity. Furthermore the Pt N edges are at a significantly smaller wavelength

compared to the Pt O

edge and will therefore yield a superior spatial resolution. Specifically,

the experimental geometry with D

FTH

¼53 mm results in a maximum recorded wave vector

transfer q

max

¼74 lm

1

at 57 eV and 97 lm

1

at 75.5 eV (cf. Equation (1)), corresponding to

encoded spatial frequencies of d

¼2p/q ¼85 nm and 65 nm, respectively. Note, that in our

experiment, the spatial resolution is also limited by the size of the reference hole and is esti-

mated to be on the order of 80 nm (see below).

The magnetic difference holograms are centered with subpixel accuracy, the sharp edge of

the beam stop is blurred by a Gaussian filter, the intensity pattern is transformed to in plane

q-coordinates

and a 2D Fourier transform yields the real-space reconstructions of the mag-

netic domain network. Finally, we interpolate the images by increasing the sampling rate by a

factor of 4. Assuming a well-defined reference wave, the real and imaginary parts of the recon-

structed images allow to deduce the dispersive and absorptive part of the dichroic index of

refraction.

However, in the XUV energy range, the wavelengths are on the order of 20 nm

and start to approach the size of the reference hole diameters such that the wave guiding effects

have to be taken into account. These are expected to exhibit a subtle dependence on the exact

shape of the reference hole and on the XUV wavelengths and, hence, will result in an addi-

tional and a priori unknown reference wave phase shift.

Therefore, we define a measure for

the total magnetic contrast as the sum of the real and imaginary part of the reconstructions. In

Figure 3, we show the resulting real space images of the magnetic domain network as a func-

tion of real space coordinates and for all recorded energies. The black and white regions within

the circular field of view correspond to areas of magnetization pointing into opposite out-of-

plane directions. The color map is scaled from the minimum to the maximum value within

each image. We observe clear and well resolved domain patterns over the entire energy range.

Note that the images have the same number of pixels in both spatial dimensions, the size of

one pixel in the reconstruction decreases from 41.7 nm at 57 eV to 31.9 nm at 75 eV (without

interpolation). The contrast is inverted for energies larger than 71 eV in agreement with meas-

urements showing an opposite sign of the MCD effect between the Co M and Pt N

resonances.

In Figure 4, we show the average peak-to-peak magnetic domain contrast within the field

of view normalized to 1 s integration time. At the Co M edge at 60 eV, we observe a pro-

nounced maximum, smaller maxima can be assigned to the N

and N

edge of Pt at 73 eV and

71.5 eV, respectively. We determine the noise level of approximately 10

cts/s (dashed line in

Fig. 4) by calculating the average peak to peak value outside the object hole. Note that even at

the extreme ends of the energy range, at 57 eV and 76 eV, we can detect high contrast domain

patterns, with signal to noise ratios exceeding 3.5 and 2, respectively. The energy dependent

contrast variations are in qualitative agreement with the small-angle scattering intensity shown

in Fig. 2(b). We attribute the quantitative differences to slightly different properties of the Co/

Pt interfaces of the imaging sample (Fig. 1(h)), which are known to sensitively influence the

magnitude of the induced magnetization of Pt. The fact that our step width of 0.5 eV undersam-

ples the spectrally narrow Pt N

edge may also cause further quantitative differences.

014301-5 Willems et al. Struct. Dyn. 4, 014301 (2017)

In Figure 5, we present a detail of the measured domain contrast at 71.5 eV and 60 eV.

The images are scaled to their actual real-space dimension in nanometer, due to the shorter

wavelength one pixel in the reconstruction at 71.5 eV corresponds to 33.6 nm and at 60 eV to

40.0 nm respectively (after interpolation, these numbers reduce to 10 nm and 8.4 nm). In

Figure 5(c), we show the normalized line profiles calculated along the white line shown in

(a) and (b) and corrected for the inverted contrast. Because in our FTH experiment, the spa-

tial resolution is determined by the reference hole geometry, rather than by the numerical

aperture of the setup and wavelength, both measurements have the same resolution on the

order of 80 nm. An exact determination of the spatial resolution based on these magnetic

images is challenging because on the one hand, a slight high pass-filtering is present due to

the use of a central beam stop and because a finite size of the domain wall width has to be

taken into account.

FIG. 3. Reconstructions of the magnetic domain network measured for XUV photon energies from 57.0 eV to 76 eV. High-

resolution real-space images of the domain network are reconstructed for the entire energy range. Note that for increasing

XUV photon energies (smaller wavelengths), the scaling of the images decreases from 42 nm/pixel at 57 eV to 32 nm/pixels

at 75 eV. The contrast at the Pt N

6,7

edges (>71 eV) is inverted.

014301-6 Willems et al. Struct. Dyn. 4, 014301 (2017)

The analysis in Figure 5(c) allows an element specific comparison of the lateral magnetiza-

tion profile with a high spatial resolution. For the investigated Co/Pt multilayer, we observe an

identical magnetic domain pattern, indicating a laterally homogeneous magnetization throughout

the entire film thickness. In heterostructures or bilayers of Co/Pt, the equilibrium magnetization

of Pt is strictly confined to the vicinity of the boundary

19,24

and the properties of the interface

magnetism are governed by spin-orbit interaction. Here, we expect an optically or electrically

induced spin injection to lead to a transient spatial rearrangement of magnetic order at the Co/

Pt boundary. This makes multi-color coherent imaging experiments a unique experimental tool

to study the element-specific response of magnetization dynamics in three-dimensional space

and promises to shed light on a wealth of intriguing Co/Pt interface phenomena.

1–9

IV. SIMULTANEOUS TWO-COLOR COHERENT IMAGING

In the following, we present the first experimental realization of a coherent imaging experi-

ment with direct and simultaneous access to the element specific and spatially resolved magne-

tization of two distinct elements, Co and Pt.

The experiment was carried out at the free electron laser (FEL) facility FERMI delivering

brilliant, femtosecond pulses in the XUV spectral region.

31,32

Briefly, FERMI relies on a seeded

harmonic scheme where the FEL emission occurs at a harmonic of an external UV seed pulse.

In a first undulator, the seed interacts with a bunch of relativistic electrons and modulates their

energy longitudinally with the periodicity of the seed wavelength, k

Seed

. Then, in a magnetic

FIG. 4. Average peak-to-peak magnetic domain contrast in counts per second as a function of XUV photon energy. We

observe a pronounced peak at the Co edge at E

¼60 eV and two smaller maxima at 74 eV and 71.5 eV which we attribute

to core-valence transitions of Pt N

6,7

. The horizontal dashed blue line shows the background signal, i.e., the peak to peak

values outside the field of view.

014301-7 Willems et al. Struct. Dyn. 4, 014301 (2017)

chicane, electrons follow an energy-dependent path which converts the energy modulation into

an electron density modulation, forming micro-bunches that emit coherently in a second undula-

tor section, tuned to the desired harmonic wavelength.

For simultaneous probing of the magnetizations of Co and Pt, a two color operation of the

FEL is required. For this purpose, the second undulator was split into two subsections, being

resonant at k

FEL,1

¼k

Seed,1

/mand k

FEL,2

¼k

Seed,1

/n, where nand mare the integer harmonic

numbers. In such a configuration, the two colors are synchronized and probe the sample simul-

taneously. More advanced generation schemes may be used for introducing a delay of few hun-

dreds of femtoseconds between the two colors,

12,13,33,34

while a XUV-split-and-delay scheme

offers a complete control over the temporal separation and spatial overlap of the probe pulses,

at the cost of a more complex experimental implementation.

The constraint for the accessible FEL wavelength separation is shown in Figure 6(a) and is

given by the photon energy of the UV seed laser or multiples of it, i.e., multiples of approxi-

mately 5 eV. We replot the energy dependent peak to peak magnetic domain contrast of Figure 3

and indicate the optimal FEL wavelengths by solid orange lines in Figures 6(a) and 6(b).Acom-

parable magnetic contrast for Pt and Co is obtained at 71.6 eV (k

FEL,2

¼17.3 nm) and 61.4 eV

FEL,1

¼20.2 nm), respectively.

For the two-color imaging experiment, we adapted the FTH mask geometry to avoid a spa-

tial overlap of the reconstructed objects. In the reconstruction of the (nn)-sized hologram,

the position of the object is wavelength dependent and is given by

xðkFELÞ¼

r=rndpixr=ðkFELDFTHÞ;(2)

in units of pixels (cf. Fig. 1(f)). The vector connecting the reference and object hole is denoted

by ~

r. Hence, spatially separated images of the dr¼2lm sized object at the two different FEL

wavelengths are achieved by adding two additional reference holes at a larger distance of

r¼j

rj¼13 lm. With such multi-reference FTH imaging,

additional care has to be taken, that

none of the cross correlations between the various reference holes overlap with the object-

reference correlation of interest.

FIG. 5. (a) Detail of the reconstructed magnetic domains (a) at 71.5 eV and (b) at 60.0 eV as a function of real space coordi-

nates. (c) Normalized lineouts along the white lines shown in (a) and (b). We observe an inverted, yet identical magnetic

domain pattern for Co and Pt layers. The spatial resolution is estimated to be below 80 nm. The solid line is a guide to the eye.

014301-8 Willems et al. Struct. Dyn. 4, 014301 (2017)

The experiment was carried out at the end-station DiProI,

37,38

the geometry of the setup as

well as the sample (cf. Figs. 1(i) and 1(h), respectively) is identical to the synchrotron measure-

ments described above. We reduce the FEL intensity of the two pulses to 2 60.3 lJ at 61.4 eV

(Co) and 3 60.3 lJ at 71.6 eV (Pt) with a spot size of 180 lm190 lm (FWHM) in order to

avoid the X-ray induced changes of the domain pattern.

16,39

This corresponds to approximately

810

photons/lm

/pulse and 10

photons/lm

/pulse at the Co and Pt resonance, respectively.

At a 10 Hz repetition rate, we acquire 5 images with 600 s integration time for both left and

right circular polarization. We repeat the same analysis for the digital image reconstruction as

described above and additionally increase the signal to noise ratio by summing the reconstruc-

tions from the two new reference holes. The simultaneously measured, element-specific, real-

space images of the magnetic domain pattern are presented in Figure 6(c). The magnetic

domains are clearly resolved, and the resolution is comparable to the measurements presented

in Fig. 3. The reconstruction shows no imaging artefacts and has an excellent suppression of

the charge scattering.

In the following, we make a conservative estimate on the number of required XUV photons

to perform a two-color imaging experiment with <80 nm resolution. Assuming a reduced object

hole size in the dispersion direction of dr¼1lm and a relative wavelength difference Dk=k

0:15 (e.g., Pt N

and Co M

2,3

edges) a spatial separation of the two reconstructed images

requires a minimum length of the vector j~

rjconnecting the reference and the object hole of

r¼drk=Dk6lmorr=dr¼6. Focusing to a spot size on the sample of 12 lm (FWHM) to

homogenously illuminate the reference and object and maintain the photon flux and integration

times of the above describe FEL experiment, we would require approximately 10

photons/s in

the two-color XUV beam. With advanced reference schemes like multi-reference geometries

or monolithic zone plate focusing reference structures

the signal noise ratio can be further sig-

nificantly improved. We thus expect that multi-color imaging experiments will be feasible in

the near future with lab-based high-harmonic sources.

We note that by increasing the ratio

r=dr20 of the FTH mask, one will even be able to use the entire high harmonic spectrum

generated by a k¼800 nm driver laser without any further wavelength selecting optics.

Spatially and temporally resolved spectroscopy with double- or multi-color XUV probe

pulses offers the unique opportunity to simultaneously access the element- or electronic-specific

FIG. 6. (a), (b) FEL energy as a function of seed wavelength shown for harmonics H12 to H15. For a single seed wave-

length of k

SEED

¼242.2 nm, we can maximize the magnetic domain contrast at the Pt N

edge at 71.6 eV (H14) and simul-

taneously get a comparable signal for H12 at 61.4 eV at the Co M

2,3

edge. (c) Element specific magnetic domain patterns

for Pt and Co reconstructed from a single difference hologram. The respective element specific real space information for

Co and Pt does not overlap. Note that the pixels correspond to different real-space coordinates due to the different XUV

wavelengths employed.

014301-9 Willems et al. Struct. Dyn. 4, 014301 (2017)

response in a single experiment. This is not only of imminent importance for non-repetitive

experiments of stochastic processes

or for very high, destructive excitation densities,

but

also for complex multi-component or multiphase systems where the excitation is followed by a

complex and ultrafast interaction between different constituent elements or different electronic

states. Some recent and prominent examples include competing phases in correlated materials

showing metal to insulator transitions

43–46

and chemical inhomogeneities in ferrimagnets exhib-

iting all-optical switching.

V. CONCLUSION

We have demonstrated spatially resolved access to element-specific magnetization in Co/Pt

heterostructures, both in reciprocal space via SAXS and in real space via FTH. The XMCD

effect at the Co M

2,3

edge as well as the very strong dichroic signals at the O

2,3

and N

6,7

edge

of Pt give rise to almost background free magnetic scattering signals and lead to bright diffrac-

tion peaks in q-space and high-contrast and high-resolution magnetic domain images in real

space. We presented the first realization of a double-color imaging experiment at the free-

electron laser facility FERMI encoding the real space magnetic domain patterns of Co and Pt

in a single hologram. We envision the multi-color, real-space spectroscopy at FEL and HHG

sources to become a valuable tool to unravel ultrafast interactions within the electronic and

spin structure of complex multicomponent and multiphase materials.

ACKNOWLEDGMENTS

The outstanding support provided during preparation and realization of the experiments by the

scientists of the different divisions of FERMI is greatly appreciated. The Max-Born-Institute group

acknowledges financial support received from the Helmholtz Virtual Institute “Dynamic Pathways

in Multidimensional Landscapes” (VH-VI-419). Support from the CNRS through the “PEPS

SASELEX” and from the French ANR via the “UMAMI” project is acknowledged by the co-

authors from Paris. B.M. acknowledges the financial support by the “ERC X-Five” grant.

K. S. Ryu, L. Thomas, S.-H. Yang, and S. S. P. Parkin, “Chiral spin torque at magnetic domain walls,” Nat. Nanotechnol.

8(7), 527–533 (2013).

S. Woo, K. Litzius, B. Kr€

uger, M.-Y. Im, L. Caretta, K. Richter, M. Mann, A. Krone, R. M. Reeve, M. Weigand, P.

Agrawal, I. Lemesh, M.-A. Mawass, P. Fischer, M. Kl€

aui, and G. S. D. Beach, “Observation of room-temperature mag-

netic skyrmions and their current-driven dynamics in ultrathin metallic ferromagnets,” Nat. Mater. 15(5), 501–506

(2016).

T. J. Huisman, R. V. Mikhaylovskiy, J. D. Costa, F. Freimuth, E. Paz, J. Ventura, P. P. Freitas, S. Bl€

ugel, Y. Mokrousov,

Th. Rasing, and A. V. Kimel, “Femtosecond control of electric currents in metallic ferromagnetic heterostructures,” Nat.

Nanotechnol. 11, 455–458 (2016).

T. Kampfrath, M. Battiato, P. Maldonado, G. Eilers, J. N€

otzold, S. M€

ahrlein, V. Zbarsky, F. Freimuth, Y. Mokrousov, S.

Bl€

ugel, M. Wolf, I. Radu, P. M. Oppeneer, and M. M€

unzenberg, “Terahertz spin current pulses controlled by magnetic

heterostructures,” Nat. Nanotechnol. 8(4), 256–260 (2013).

T. Seifert, S. Jaiswal, U. Martens, J. Hannegan, L. Braun, P. Maldonado, F. Freimuth, A. Kronenberg, J. Henrizi, I. Radu,

E. Beaurepaire, Y. Mokrousov, P. M. Oppeneer, M. Jourdan, G. Jakob, D. Turchinovich, L. M. Hayden, M. Wolf, M.

M€

unzenberg, M. Kl€

aui, and T. Kampfrath, “Efficient metallic spintronic emitters of ultrabroadband terahertz radiation,”

Nat. Photonics 10, 483–488 (2016).

C.-H. Lambert, S. Mangin, B. S. D. Ch. S. Varaprasad, Y. K. Takahashi, M. Hehn, M. Cinchetti, G. Malinowski, K.

Hono, Y. Fainman, M. Aeschlimann, and E. E. Fullerton, “All-optical control of ferromagnetic thin films and nano-

structures,” Science 345(6202), 1337–1340 (2014).

M. O. A. Ellis, E. E. Fullerton, and R. W. Chantrell, “All-optical switching in granular ferromagnets caused by magnetic

circular dichroism,” Sci. Rep. 6, 30522 (2016).

R. John, M. Berritta, D. Hinzke, C. M€

uller, T. Santos, H. Ulrichs, and P. Nieves, “Magnetization switching of FePt nano-

particle recording medium by femtosecond laser pulses,” e-print arXiv:1606.08723 (2016).

Y. K. Takahashi, R. Medapalli, S. Kasai, J. Wang, K. Ishioka, S. H. Wee, O. Hellwig, K. Hono, and E. E. Fullerton,

“Accumulative magnetic switching of ultra-high-density recording media by circularly polarized light,” Phys. Rev. Appl.

6, 054004 (2016).

R. Medapalli, D. Afanasiev, D. K. Kim, Y. Quessab, S. Manna, S. A. Montoya, A. Kirilyuk, Th. Rasing, A. V. Kimel,

and E. E. Fullerton, “Mechanism of all-optical control of ferromagnetic multilayers with circularly polarized light,” e-

print arXiv:1607.02505 (2016).

M. S. El Hadri, M. Hehn, P. Pirro, C.-H. Lambert, G. Malinowski, E. E. Fullerton, and S. Mangin, “Domain size criterion

for the observation of all-optical helicity-dependent switching in magnetic thin films,” Phys. Rev. B 94(6), 064419

(2016).

014301-10 Willems et al. Struct. Dyn. 4, 014301 (2017)

E. Allaria, F. Bencivenga, R. Borghes, F. Capotondi, D. Castronovo, P. Charalambous, P. Cinquegrana, M. B. Danailov,

G. De Ninno, A. Demidovich, S. Di. Mitri, B. Diviacco, D. Fausti, W. M. Fawley, E. Ferrari, L. Froehlich, D. Gauthier,

A. Gessini, L. Giannessi, R. Ivanov, M. Kiskinova, G. Kurdi, B. Mahieu, N. Mahne, I. Nikolov, C. Masciovecchio, E.

Pedersoli, G. Penco, L. Raimondi, C. Serpico, P. Sigalotti, S. Spampinati, C. Spezzani, C. Svetina, M. Trov

o, and M.

Zangrando, “Two-colour pump-probe experiments with a twin-pulse-seed extreme ultraviolet free-electron laser,” Nat.

Commun. 4, 2476 (2013).

B. Mahieu, E. Allaria, and D. Castronovo, “Two-colour generation in a chirped seeded free-electron laser: A close look,”

Opt. Express 21(19), 113–119 (2013).

G. Schiwietz, M. Beye, and T. Kachel, “UE112_ PGM-1: An open-port low-energy beamline at the BESSY II undulator

UE112,” J. Large-Scale Res. Facil. 1, A33 (2015).

O. Hellwig, G. P. Denbeaux, J. B. Kortright, and Eric E. Fullerton, “X-ray studies of aligned magnetic stripe domains in

perpendicular multilayers,” Phys. B: Condens. Matter 336(1–2), 136–144 (2003).

C. Gutt, S. Streit-Nierobisch, L.-M. Stadler, B. Pfau, C. M. G€

unther, R. K€

onnecke, R. Fr€

omter, A. Kobs, D. Stickler, H.

P. Oepen, R. R. F€

austlin, R. Treusch, J. Feldhaus, E. Weckert, I. A. Vartanyants, M. Grunze, A. Rosenhahn, T. Wilhein,

S. Eisebitt, and G. Gr€

ubel, “Single-pulse resonant magnetic scattering using a soft x-ray free-electron laser,” Phys. Rev.

B81(10), 100401 (2010).

B.Pfau,S.Schaffert,L.M

€

uller,C.Gutt,A.Al-Shemmary,F.B

€

uttner,R.Delaunay,S.D

€

usterer, S. Flewett, R. Fr€

omter, J.

Geilhufe, E. Guehrs, C. M. G€

unther, R. Hawaldar, M. Hille, N. Jaouen, A. Kobs, K. Li, J. Mohanty, H. Redlin, W. F.

Schlotter, D. Stickler, R. Treusch, B. Vodungbo, M. Kl€

aui, H. P. Oepen, J. L€

uning, G. Gr€

ubel, and S. Eisebitt, “Ultrafast opti-

cal demagnetization manipulates nanoscale spin structure in domain walls,” Nat. Commun. 3, 1100 (2012).

M. Schneider, C. M. G€

unther, C. von Korff Schmising, B. Pfau, and S. Eisebitt, “Curved gratings as an integrated

photon fluence monitor in x-ray transmission scattering experiments,” Opt. Express 24(12), 13091 (2016).

N. Nakajima, T. Koide, T. Shidara, H. Miyauchi, H. Fukutani, A. Fujimori, K. Iio, T. Katayama, M. N

yvlt, and Y.

Suzuki, “Perpendicular magnetic anisotropy caused by interfacial hybridization via enhanced orbital moment in Co/Pt

multilayers: Magnetic circular x-ray dichroism study,” Phys. Rev. Lett. 81(23), 5229–5232 (1998).

T. Shishidou, S. Imada, T. Muro, F. Oda, A. Kimura, and S. Suga, “Strong Fano effect in the magnetic circular dichroism

of the Pt N

6,7

core absorption of ferromagnetic CoPt

,” Phys. Rev. B 55(6), 3749 (1997).

F. Willems, C. T. L. Smeenk, N. Zhavoronkov, O. Kornilov, I. Radu, M. Schmidbauer, M. Hanke, C. von Korff

Schmising, M. J. J. Vrakking, and S. Eisebitt, “Probing ultrafast spin dynamics with high-harmonic magnetic circular

dichroism spectroscopy,” Phys. Rev. B - Condens. Matter Mater. Phys. 92, 220405 (2015).

J. B. Kortright and S.-K. Kim, “Resonant magneto-optical properties of Fe near its 2 p levels,” Phys. Rev. B 62(18),

216–228 (2000).

B. Vodungbo, J. Gautier, G. Lambert, A. Barszczak Sardinha, M. Lozano, S. Sebban, M. Ducousso, W. Boutu, K. Li, B.

Tudu, M. Tortarolo, R. Hawaldar, R. Delaunay, V. L

opez-Flores, J. Arabski, C. Boeglin, H. Merdji, P. Zeitoun, and J.

L€

uning, “Laser-induced ultrafast demagnetization in the presence of a nanoscale magnetic domain network,” Nat.

Commun. 3, 999 (2012).

J. Geissler, E. Goering, M. Justen, F. Weigand, G. Sch€

utz, J. Langer, D. Schmitz, H. Maletta, and R. Mattheis, “Pt magne-

tization profile in a Pt/Co bilayer studied by resonant magnetic x-ray reflectometry,” Phys. Rev. B 65, 020405 (2001).

D. Paterson, B. E. Allman, P. J. McMahon, J. Lin, N. Moldovan, K. A. Nugent, I. McNulty, C. T. Chantler, C. C. Retsch,

T. H. K. Irving, and D. C. Mancini, “Spatial coherence measurement of X-ray undulator radiation,” Opt. Commun.

195(1–4), 79–84 (2001).

B. Pfau and S. Eisebitt, “X-Ray holography,” in Synchrotron Light Sources and Free-Electron Lasers, edited by E. J.

Jaeschke, S. Khan, J. R. Schneider, and J. B. Hastings (Springer International Publishing, 2016), pp. 1093–1133.

S. Eisebitt, J. L€

uning, W. F. Schlotter, O. Hellwig, W. Eberhardt, and J. St€

ohr, “Lensless imaging of magnetic nanostruc-

tures by X-ray spectro-holography,” Nature 432, 885–888 (2004).

S. Schaffert, B. Pfau, J. Geilhufe, C. M. G€

unther, M. Schneider, C. von Korff Schmising, and S. Eisebitt, “High-resolu-

tion magnetic-domain imaging by Fourier transform holography at 21 nm wavelength,” New J. Phys. 15, 093042 (2013).

A. Scherz, W. Schlotter, K. Chen, R. Rick, J. St€

ohr, J. L€

uning, I. McNulty, C. M. G€

unther, F. Radu, W. Eberhardt, O.

Hellwig, and S. Eisebitt, “Phase imaging of magnetic nanostructures using resonant soft x-ray holography,” Phys. Rev. B

76, 214410 (2007).

S. Zayko, E. M€

onnich, M. Sivis, D.-D. Mai, T. Salditt, S. Sch€

afer, and C. Ropers, “Coherent diffractive imaging beyond

the projection approximation: Waveguiding at extreme ultraviolet wavelengths,” Opt. Express 23(15), 19911 (2015).

G. De Ninno, D. Gauthier, B. Mahieu, P. R. Ribic, E. Allaria, P. Cinquegrana, M. B. Danailov, A. Demidovich, E.

Ferrari, L. Giannessi, G. Penco, P. Sigalotti, and M. Stupar, “Single-shot spectro-temporal characterization of XUV

pulses from a seeded free-electron laser,” Nat. Commun. 6, 8075 (2015).

B. Mahieu, D. Gauthier, G. De Ninno, H. Dacasa, M. Lozano, J.-P. Rousseau, P. Zeitoun, D. Garzella, and H. Merdji,

“Spectral-phase interferometry for direct electric-field reconstruction applied to seeded extreme-ultraviolet free-electron

lasers,” Opt. Express 23(14), 17665–17674 (2015).

E. Ferrari, C. Spezzani, F. Fortuna, R. Delaunay, F. Vidal, I. Nikolov, P. Cinquegrana, B. Diviacco, D. Gauthier, G.

Penco, P. R. Ribic, E. Roussel, M. Trov

o, J.-B. Moussy, T. Pincelli, L. Lounis, M. Manfredda, E. Pedersoli, F.

Capotondi, C. Svetina, N. Mahne, M. Zangrando, L. Raimondi, A. Demidovich, L. Giannessi, G. De Ninno, M. B.

Danailov, E. Allaria, and M. Sacchi, “Widely tunable two-colour seeded free-electron laser source for resonant-pump res-

onant-probe magnetic scattering,” Nat. Commun. 7, 10343 (2016).

D. Gauthier, P. R. Ribic, Gi. De Ninno, E. Allaria, P. Cinquegrana, M. B. Danailov, A. Demidovich, E. Ferrari, L.

Giannessi, B. Mahieu, and G. Penco, “Spectrotemporal shaping of seeded free-electron laser pulses,” Phys. Rev. Lett.

115, 114801 (2015).

B. Pfau, C. M. G€

unther, S. Schaffert, R. Mitzner, B. Siemer, S. Roling, H. Zacharias, O. Kutz, I. Rudolph, R. Treusch,

and S. Eisebitt, “Femtosecond pulse X-ray imaging with a large field of view,” New J. Phys. 12, 095006 (2010).

W. F. Schlotter, R. Rick, K. Chen, A. Scherz, J. St€

ohr, J. L€

uning, S. Eisebitt, C. M. G€

unther, W. Eberhardt, O. Hellwig,

and I. McNulty, “Multiple reference Fourier transform holography with soft x rays,” Appl. Phys. Lett. 89(16), 2004–2007

(2006).

014301-11 Willems et al. Struct. Dyn. 4, 014301 (2017)

F. Capotondi, E. Pedersoli, N. Mahne, R. H. Menk, G. Passos, F. Capotondi, E. Pedersoli, N. Mahne, R. H. Menk, G.

Passos, and L. Raimondi, “Invited Article: Coherent imaging using seeded free-electron laser pulses with variable polari-

zation: First results and research opportunities,” Rev. Sci. Instrum. 84, 051301 (2013).

C. von Korff Schmising, B. Pfau, M. Schneider, C. M. G€

unther, M. Giovannella, J. Perron, B. Vodungbo, L. M€

uller, F.

Capotondi, E. Pedersoli, N. Mahne, J. L€

uning, and S. Eisebitt, “Imaging ultrafast demagnetization dynamics after a spa-

tially localized optical excitation,” Phys. Rev. Lett. 112, 217203 (2014).

T. Wang, D. Zhu, B. Wu, C. Graves, S. Schaffert, T. Rander, L. M€

uller, B. Vodungbo, C. Baumier, D. P. Bernstein, B.

Br€

auer, V. Cros, S. De Jong, R. Delaunay, A. Fognini, R. Kukreja, S. Lee, V. L

opez-Flores, J. Mohanty, B. Pfau, H.

Popescu, M. Sacchi, A. B. Sardinha, F. Sirotti, P. Zeitoun, M. Messerschmidt, J. J. Turner, W. F. Schlotter, O. Hellwig,

R. Mattana, N. Jaouen, F. Fortuna, Y. Acremann, C. Gutt, H. A. D€

urr, E. Beaurepaire, C. Boeglin, S. Eisebitt, G. Gr€

ubel,

J. L€

uning, J. St€

ohr, and A. O. Scherz, “Femtosecond single-shot imaging of nanoscale ferromagnetic order in Co/Pd mul-

tilayers using resonant x-ray holography,” Phys. Rev. Lett. 108, 267403 (2012).

J. Geilhufe, B. Pfau, M. Schneider, F. B€

uttner, C. M. G€

unther, S. Werner, S. Schaffert, E. Guehrs, S. Fr€

ommel, M. Kl€

aui,

and S. Eisebitt, “Monolithic focused reference beam X-ray holography,” Nat. Commun. 5, 3008 (2014).

G. O. Williams, A. I. Gonzalez, S. K€

unzel, L. Li, M. Lozano, E. Oliva, B. Iwan, S. Daboussi, W. Boutu, H. Merdji, M.

Fajardo, and P. Zeitoun, “Fourier transform holography with high harmonic spectra for attosecond imaging applications,”

Opt. Lett. 40(13), 3205–3208 (2015).

L. M€

uller, C. Gutt, B. Pfau, S. Schaffert, J. Geilhufe, F. B€

uttner, J. Mohanty, S. Flewett, R. Treusch, S. D€

usterer, H.

Redlin, A. Al-Shemmary, M. Hille, A. Kobs, R. Fr€

omter, H. P. Oepen, B. Ziaja, N. Medvedev, S. K. Son, R. Thiele, R.

Santra, B. Vodungbo, J. L€

uning, S. Eisebitt, and G. Gr€

ubel, “Breakdown of the X-Ray resonant magnetic scattering signal

during intense pulses of extreme ultraviolet free-electron-laser radiation,” Phys. Rev. Lett. 110, 234801 (2013).

M. H. Burkhardt, M. A. Hossain, S. Sarkar, Y. D. Chuang, A. G. Cruzgonzalez, A. Doran, A. Scholl, A. T. Young, N.

Tahir, Y. J. Choi, S. W. Cheong, H. A. D€

urr, and J. St€

ohr, “Imaging the first-order magnetic transition in

0.35

0.275

0.375

MnO

,” Phys. Rev. Lett. 108, 237202 (2012).

S. de Jong, R. Kukreja, C. Trabant, N. Pontius, C. F. Chang, T. Kachel, M. Beye, F. Sorgenfrei, C. H. Back, B. Br€

auer,

W. F. Schlotter, J. J. Turner, O. Krupin, M. Doehler, D. Zhu, M. A. Hossain, A. O. Scherz, D. Fausti, F. Novelli, M.

Esposito, W. S. Lee, Y. D. Chuang, D. H. Lu, R. G. Moore, M. Yi, M. Trigo, P. Kirchmann, L. Pathey, M. S. Golden, M.

Buchholz, P. Metcalf, F. Parmigiani, W. Wurth, A. F€

ohlisch, C. Sch€

ußler-Langeheine, and H. A. D€

urr, “Speed limit of

the insulator-metal transition in magnetite,” Nat. Mater. 12(10), 882–886 (2013).

S. A. D€

onges, O. Khatib, B. T. O’Callahan, J. M. Atkin, J. Hyung Park, D. Cobden, and M. B. Raschke, “Ultrafast nanoi-

maging of the photoinduced phase transition dynamics in VO

,” Nano Lett. 16, 3029–3035 (2016).

L. Vidas, C. M. G€

unther, T. A. Miller, B. Pfau, M. Schneider, E. Guehrs, R. E. Marvel, K. A. Hallman, R. F. Haglund,

Jr., S. Eisebitt, and S. Wall, “Nanoscale spectroscopic imaging of phase separation in a correlated material by resonant

X-ray holography,” e-print arXiv:1612.07998 (2016).

C. E. Graves, A. H. Reid, T. Wang, B. Wu, S. de Jong, K. Vahaplar, I. Radu, D. P. Bernstein, M. Messerschmidt, L.

M€

uller, R. Coffee, M. Bionta, S. W. Epp, R. Hartmann, N. Kimmel, G. Hauser, A. Hartmann, P. Holl, H. Gorke, J. H.

Mentink, A. Tsukamoto, A. Fognini, J. J. Turner, W. F. Schlotter, D. Rolles, H. Soltau, L. Str€

uder, Y. Acremann, A. V.

Kimel, A. Kirilyuk, Th. Rasing, J. St€

ohr, A. O. Scherz, and H. A. D€

urr, “Nanoscale spin reversal by non-local angular

momentum transfer following ultrafast laser excitation in ferrimagnetic GdFeCo,” Nat. Mater. 12(4), 293–298 (2013).

014301-12 Willems et al. Struct. Dyn. 4, 014301 (2017)