Spin Wave Dispersion Relation Engineering by Magnonic Crystals With Arbitrary Symmetry
Spin Wave Dispersion Relation Engineering by Magnonic Crystals With Arbitrary Symmetry
Yesenia Angélica García Jomaso, David Ley Dominguez, Francisco Javier Peña Espinoza, et al.
COLLECTIONS
© 2023 Author(s).
Journal of ARTICLE [Link]/journal/jap
Applied Physics
Yesenia Angélica García Jomaso,1 David Ley Dominguez,1 Francisco Javier Peña Espinoza,1
1 1
Elizabeth Mendoza Sandoval, Arturo Reyes Almanza, Naser Qureshi,2
1
César Leonardo Ordóñez-Romero, and Giuseppe Pirruccio1,a)
AFFILIATIONS
1
Instituto de Física, Universidad Nacional Autónoma de México, Apartado Postal 20-364, Ciudad de México C.P. 01000, Mexico
2
ICAT, Universidad Nacional Autónoma de México, CU, México CDMX 04510, Mexico
a)
Author to whom correspondence should be addressed: pirruccio@[Link]
ABSTRACT
The use of metasurfaces to engineer the response of magnetic materials is of utmost importance in the field of magnon-spintronics.
Here, we demonstrate a method to fabricate one- and two-dimensional magnonic crystals with arbitrary symmetry and use it to engineer
the amplitude-frequency characteristic of magnetostatic surface spin waves excited in a magnetic material. The technique is based on the
gentle microablation of the sample surface by focused femtosecond laser pulses. Tightly focused illumination allows using modest pulse
energy while achieving micrometer precision. By raster scanning the incident laser spot on the sample surface, we control the shape and
size of the building blocks constituting the unit cell of the crystal along with its symmetry and lattice parameter. Remarkable and controlled
changes in the measured transmission characteristics reveal the strong and complex symmetry-dependent interaction of the spin waves with
Bravais and non-Bravais lattices. The described single-step microfabrication method facilitates and speeds up the realization of integrated
spintronics components and provides an efficient tool to explore complex magnetic dynamics in scattering lattices.
departs from the ideal profile of the mask. A review of other tech-
niques used to shape the magnonic response of magnetic materi-
als is found in Ref. 21. The highest spatial resolution is achieved
with electron beam lithography22 and focused ion beam,23 which
despite permitting nanometer precision are affected by many of
the setbacks already mentioned, need a clean room, is extremely
hard to fabricate mm-size structures, and are slow and expensive.
Here, we propose and demonstrate a fast, reliable, and rela-
tively inexpensive, single-step microfabrication method to realize
MCs of arbitrary symmetry on cm2 areas maintaining a micrometer
precision. We use femtosecond laser pulses at a high repetition rate
tightly focused by an optical objective to microablate the surface of
a slab of yttrium iron garnet (YIG). In order to control the size
and shape of the building blocks composing the unit cell as well as
the lattice symmetry, the laser shutter and stepper motors are syn-
chronized. Advanced geometries often determine the observation
of unexpected phenomena and physical properties. For this reason,
we pattern the surface with one-dimensional and two-dimensional
FIG. 1. Sketch of the experimental setup used to fabricate the MCs and
MCs and measure their amplitude-frequency characteristic (AFC). measure their AFC.
Drastic changes in the transmission curve and the appearance of
several bandgaps reveal the resonant interaction of the surface spin
waves with the different lattices. Using a fluence close to the abla-
tion threshold permits us to obtain high-quality features and a 2 μm-spot, measured at its full width at half maximum, using a
observe fine details in the measured spectra. We explore the intro- 0.25 numerical aperture microscope objective (Olympus, plan ach-
duced complexity of the electromagnetic interactions by comparing romat, 10X). Although the spot size is not truly diffraction-limited,
the response of Bravais and non-Bravais lattices, where we reveal the obtained focus profile is enough to ensure micrometer precision
the effect of the multi-element motif of the unit cell. The use of during the ablation process. This is at least two orders of magni-
two-dimensional lattices implies the possible built-in conversion tude smaller than the spin wave wavelength and one order of mag-
between different types of spin waves within a single device, an nitude finer than the spatial resolution obtainable with a traditional
element that has remained unexplored so far and deserves further mask-assisted chemical etching, allowing for fine features on the
consideration. patterned area. Another important point to be considered relates to
the depth of focus. A tighter focus can be obtained by using an
objective with a larger numerical aperture or by using higher har-
II. SAMPLES AND FABRICATION METHOD monics of the fundamental laser emission. However, both solutions
Our samples consist of uniform thin YIG films epitaxially are unnecessary and expensive. In fact, a smaller diffraction-limited
grown on a gallium gadolinium garnet (GGG) substrate, along the spot provides not only a larger lateral resolution, but also a much
(111) crystallographic plane. All samples are approximately smaller penetration depth into the YIG. This, in turn, makes the
35 mm-long, 2 mm-wide, and 5 μm-thick and present a 5 GHz fer- microablation system very sensitive to small misalignment of the
romagnetic resonance half-power linewidth of about 0.5 Oe. The sample, leading to defocusing and inhomogeneities in the ablated
experimental setup used to measure the AFC is sketched in Fig. 1. depth upon desirable centimeter-long displacements. To achieve a
It is based on a spin wave delay line structure consisting of a pair reliable and stable sample movement, we employ two stepper
of 20 μm-diameter copper wire microwave antennas placed directly motors (Thorlabs, ZST225B controlled by a homemade driver)
on the surface of the YIG films. The separation between the anten- running at a constant speed of 144 μm/s. A homemade script and a
nas is chosen to be 1 cm so that the micropatterned area entirely data acquisition system trigger the laser shutter and synchronize it
fits between them. The AFC of the different MCs is measured by with the scanning stepper motors. A micrometer operated
using a vector network analyzer connected directly to the spin wave tip-tilt-rotation stage ensures the perpendicularity of the sample
delay line structure. To ensure the magnetostatic surface wave surface to the incident laser beam and its parallelism with respect
(MSSW) configuration, H is applied along the y direction parallel to the axis of one of the motors.
to the surface of the YIG film with a nominal value of 770 Oe for We first characterize the laser ablation process as a function of
all the measurements, while spin waves propagation direction is the laser parameters. The YIG ablation threshold is found by
along x. varying the pulse energy and without moving the sample. The
The MCs are fabricated by laser ablation of the YIG surface. focus diameter and the repetition rate are kept constant to approxi-
230 femtosecond (fs) pulses are generated by a high power ultrafast mately 2 μm and 1 MHz, respectively, in all the experiments. The
laser (Pharos, Light Conversion) with a repetition rate of 1 MHz. ablation process for fs pulses is found to start around 202 mJ/cm2 ,
Its fundamental emission wavelength, centered around well below the value found for ns-pulses at 1064 nm.24 It is worth
λ ¼ 1030 nm, is frequency doubled so as to obtain pulses centered noting that this value is obtained by looking in situ at the sample
at λ ¼ 515 nm. This second harmonic emission is focused down to surface during irradiation with an optical microscope without
moving the focal spot and it corresponds to the pulse energy at resolution. The top view of typical rectangular groove and squares
which the first surface damage occurs. Nevertheless, when the is displayed in Figs. 2(c) and 2(d), where the precision of the fabri-
motors scan the sample to fabricate the microstructures, the defini- cated structures can be appreciated on a larger scale. We emphasize
tion of fluence becomes somewhat ambiguous and unpractical. that no damage is observed in the proximity of the ablated area,
For this reason, we prefer to provide the incident average power confirming the precision and gentleness of the fabrication
and the speed of the motor (which we always keep constant in all method.20 Compared with visible ns-irradiation, ablation by ultra-
the experiments) for the characterization of the groove depth. In fast pulses strongly limits the material damage caused by heat while
Fig. 2(a), we determine the dependence of the ablated depth of maintaining a similar ablation threshold.25
several grooves on the incident average power. This is measured by
observing their cross section with a scanning electron microscope
(SEM). Because the speed of the sample positioner is common to III. MEASUREMENTS AND RESULTS
all the grooves, their ablated depths can be compared. A quasi- To demonstrate control over the AFC of surface spin waves,
linear regime is found, which allows us to choose the optimal we first fabricate one-dimensional MCs. We keep a constant
average power needed for a specified ablation depth. In Fig. 2(b), average power of 17.6 mW and vary the lattice parameter. In
the AFM image of a 20 μm-wide and 3.7 μm-deep groove fabricated Figs. 3(a)–3(c), optical microscope images of three MCs with
with an average power of 23.4 mW is shown. The resultant surface 70 μm-wide grooves and lattice parameter equal to 330, 430, and
has a roughness of around 230 nm, which depends on a number of 530 μm, respectively, are shown. The corresponding AFC are
factors such as the profile of the focal spot, the overlap between the plotted with blue curves in Figs. 3(d)–3(f ). We see the appearance
focal spots upon sample translation, and the stability of the motors. of several dips at specific frequencies of the S21 spectrum corre-
The groove’s trapezoidal shape is typically encountered in all fabri- sponding to the excitation of the MC Bloch modes.3,7,15 Being the
cated structures. The angles between the sides and the major wavelength of the surface spin waves comparable with the lattice
(minor) base are equal to 60 (120 ) (measured by SEM), leading parameter, these bandgaps can be interpreted in terms of diffrac-
to a half-difference between the base lengths of 2 μm. This value tion of the excited waves and, thus, their spectral position and
perfectly agrees with the spot size and can be taken as lateral number can be geometrically controlled. As the lattice parameter
FIG. 2. (a) Groove depth as a function of the laser average power measured before the microscope objective. (b) AFM image of the surface of the groove fabricated using
an average power of 23.4 mW. (c) and (d) SEM top view of a typical rectangular groove and a square.
increases, the bandgaps shift to lower frequency and new high designed and studied experimentally, all of them being readily
order bandgaps appear at higher frequencies. The chosen groove achievable with our microfabrication and measurement methods.
width maximizes the suppression of the spin wave propagation in The simplest two-dimensional MC is the simple square lattice
correspondence with the bandgaps. For the sake of completeness, shown in Fig. 4(a). Inspired by the results shown in Fig. 3(d), we
in Fig. 3(d), the reference spectrum of the unstructured YIG is chose the lattice parameter to be 330 μm along x and y and the
plotted with a red curve. We notice a relatively small attenuation of side of the square blocks forming the unit cell to be 70 μm. The
the transmission within the allowed band by about 15 dB, which AFC spectrum shows a complex response with a large number of
increases toward high frequencies. Despite this, the overall perfor- dips almost equally spaced and with different intensities. The most
mance of the microablated MCs is not compromised. The average remarkable feature appears close to 4.2 GHz where a very sharp
power used for the fabrication is a trade-off that minimizes the fab- and pronounced doublet strongly suppresses the spin wave
rication time while maintaining a high quality of the AFCs, as com- transmission.
pared with similar MCs obtained by chemical etching. Even though it is tantalizing to associate part of the features
The surface spin wave phenomenology can be enriched by shown by the AFC of the cubic MC [Fig. 4(d)] with those of the
making the waves interact with a two-dimensional potential. AFC of the one-dimensional MC [Fig. 3(d)], two things must be
A variety of Bravais and non-Bravais lattice symmetries can be noticed: (i) the interaction of spin waves with a two-dimensional
FIG. 3. (a)–(c) Optical microscope images of one-dimensional MCs having lattice parameters equal to 330, 430, and 530 μm, respectively. (d)–(f ) Blue curves correspond
to the AFCs of the one-dimensional MCs shown in (a)–(c), respectively. The red curve in (a) is the reference spectrum measured on the first MC before patterning it.
MC cannot be easily decomposed in the superposition of the symmetry trajectory the spin wave travels, i.e., the direction in
response along x and y, independently. This is evident by compar- reciprocal space parallel to the incident spin wave wavevector. We
ing the position and number of dips in Fig. 4(d) with those of fabricate a square lattice with a Wigner-Seits unit cell rotated by
Fig. 3(d) and by considering the vectorial character of the spin 45 and with a lattice parameter reduced by half with respect to
waves. (ii) Because of the direction of the applied external magnetic that shown in Fig. 4(a). The size of the building blocks is kept to
field, surface spin waves are allowed to propagate only along the x 70 μm. Therefore, Figs. 4(a) and 4(b) correspond to Γ-X and Γ-M
direction. Therefore, scattering of the spin waves at each block, high symmetry trajectory, respectively. The measured AFC plotted
which may involve a change in the wavevector, implies a conversion in Fig. 4(e) exhibits two main dips around 4.1 GHz, while the high
of surface waves into volume waves, not necessarily allowed for the frequency part of the spectrum does not seem to be affected by the
intensity and direction of the external magnetic field used in our spatial modulation introduced by the MC. A straightforward
experiment. Although propagation of volume waves in a similar increase in the complexity of the unit cell of the square lattice
structure has been observed in earlier studies.26 shown in Fig. 4(b) consists in changing the size of one of the two
The interaction of spin waves with magnonic crystals depends building blocks in its primitive unit cell, creating a two-antidot
not only on the crystal symmetry but also along which high basis square lattice. In Figs. 4(c) and 4(f), we show an optical
FIG. 4. (a)–(c) Optical microscope images of two-dimensional MCs having simple square, rotated square, and two-antidot basis square lattice symmetry, respectively.
(d)–(f ) Blue curves correspond to the AFCs of the two-dimensional MCs shown in (a)–(c), respectively. The red curve in (a) is the reference spectrum measured on the
first MC before patterning it.
microscope image of such a MC and its measured AFC, respectively. influenced by properly designed two-dimensional MCs. The
The lattice parameter is kept the same as in Fig. 4(a), but the two improvement in the surface characteristics of microstructured spin
building blocks of the unit cell measure 70 and 35 μm. By comparing waveguides introduced by our fabrication method, especially edge
the spectra in Figs. 4(e) and 4(f), we see that the two dips around smoothness, is directly related to improved physical properties.
4.1 GHz are maintained and an extra transmission dip appears at Besides the fundamental interest, all these phenomena may lead to
4.2 GHz. It is possible to hypothesize that the latter feature appears the design of more efficient spintronics devices for high-frequency
as a consequence of the different interactions of the spin waves with signal processing.
the two sublattices composing the MC. To deepen the understanding
of the behavior of spin waves in complex potentials and advancing ACKNOWLEDGMENTS
the research in magnonic crystals of arbitrary shapes, it would be
This work was supported by UNAM DGAPA PAPIIT (Grant
important to employ theoretical tools, such as micromagnetic simu-
Nos. IN104522 and IG100521) and CONACyT Project (Grant Nos.
lations, capable to account for the truly two-dimensional problem
1564464 and 1098652). The authors would like to acknowledge the
presented here and fully capture the vectorial nature of spin waves.
support provided by Arturo Rodríguez Gómez, Samuel
For example, how much defects, shape inclusions, and lattice sym-
Tehuacanero Cuapa, and Juan Gabriel Morales Morales for techni-
metry affect the spin wave spectrum in non-diffractive magnetic
cal assistance in sample preparation for electron and atomic force
systems can be quantified, as has been done in Refs. 27 and 28.
microscopies, micrographing, and analysis.
Finally, we stress that because of the geometrical limitations
imposed by the YIG waveguides and in order to maintain a lattice
AUTHOR DECLARATIONS
parameter comparable with the spin wave wavelength, the number
of lattice periods along y is necessarily smaller than that along x. In Conflict of Interest
all the fabricated MCs, 12 periods along the spin wave propagation The authors have no conflicts to disclose.
direction approximates an effectively infinite lattice.3,19,29 This need
not be the case along the y direction, which further complicates the Author Contributions
interpretation and modeling of the spectra.
Yesenia Angélica García Jomaso: Data curation (lead); Formal
analysis (equal); Investigation (equal); Methodology (supporting);
IV. CONCLUSIONS Software (lead). David Ley Dominguez: Formal analysis (equal);
In conclusion, we have demonstrated the possibility of Investigation (equal); Methodology (equal); Visualization (equal);
shaping the AFC and, thus, the dispersion relation of magnetostatic Francisco Javier Peña Espinoza: Investigation (equal);
surface spin waves by means of MCs of different symmetry in one Visualization (equal). Elizabeth Mendoza Sandoval: Formal anal-
and two dimensions. The high-quality MCs were fabricated by ysis (equal); Methodology (equal); Validation (supporting).
laser microablation with lateral and vertical micrometer precision. Arturo Reyes Almanza: Investigation (equal); Software (equal);
All the geometrical properties of the lattices can be controlled Validation (equal). Naser Qureshi: Supervision (equal);
through the fluence, repetition rate, and synchronization between Validation (equal); Writing – original draft (supporting);
laser shutter and sample micropositioner, making our fabrication Writing – review & editing (supporting). César Leonardo
method simpler, more reliable, and more precise than the com- Ordóñez-Romero: Conceptualization (equal); Funding acquisition
monly used ones for obtaining magnonic crystals of arbitrary sym- (equal); Methodology (equal); Supervision (equal); Validation
metry. Our highest quality MCs are obtained for a fluence close to (equal); Writing – original draft (equal); Writing – review &
the ablation threshold of the YIG slab and high pulse repetition editing (equal). Giuseppe Pirruccio: Conceptualization (lead);
rate. Fine-tuning the geometrical properties of the lattice allows Funding acquisition (lead); Investigation (equal); Methodology
detailed studies of the interaction of surface spin waves with reso- (equal); Project administration (lead); Resources (equal);
nant structures of complex geometry. As a final note, we emphasize Supervision (equal); Validation (equal); Writing – original draft
that our fabrication method is general; thus, it can be applied to (lead); Writing – review & editing (lead).
realize circuitry elements based on graded index magnonics, such
as focusing30 and steering lenses.31 This represents an alternative DATA AVAILABILITY
method with respect to those based on laser irradiation and local The data that support the findings of this study are available
temperature changes.32,33 Our work paves the way for the realiza- within the article.
tion of analogies of exotic phenomena observed in other physical
realms, such as topological states, backscattering-free, and unidirec- REFERENCES
tional propagation, non-Hermitian physics, magnon condensation
in lattices, Bloch oscillation, energy localization, as well as the
1
B. Lenk, H. Ulrichs, F. Garbs, and M. Münzenberg, “The building blocks of
magnonics,” Phys. Rep. 507, 107–136 (2011).
recently proposed magnon valleytronics,34 where the exquisite 2
A. V. Chumak, A. A. Serga, and B. Hillebrands, “Magnonic crystals for data
control of the phase of the scattered waves and the system dimen- processing,” J. Phys. D: Appl. Phys. 50, 244001 (2017).
sionality play a crucial role. The higher dimensionality provides 3
C. L. Ordóñez-Romero, Z. Lazcano-Ortiz, G. Pirruccio, A. Drozdovskii,
unprecedent flexibility with respect to the more traditional one- B. Kalinikos, M. Urbanek, M. O. Vigueras-Zúñiga, D. Matatagui Cruz,
dimensional magnonic crystals. Spin wave temporal dynamics, as N. Qureshi, O. Kolokoltsev, and G. Monsivais, “Pulsed spin wave propagation in
well as non-linear magnonics, are also expected to be strongly a magnonic crystal,” J. Appl. Phys. 126, 083902 (2019).
4
A. V. Chumak, A. A. Serga, and B. Hillebrands, “Magnon transistor for all- 20
C. L. Ordóñez-Romero, Z. Lazcano-Ortiz, A. Drozdovskii, B. Kalinikos,
magnon data processing,” Nat. Commun. 5, 4700 (2014). M. Aguilar-Huerta, J. L. Domínguez-Juárez, G. Lopez-Maldonado, N. Qureshi,
5
F. Busse, M. Mansurova, B. Lenk, M. von der Ehe, and M. Münzenberg, “A sce- O. Kolokoltsev, and G. Monsivais, “Mapping of spin wave propagation in a one-
nario for magnonic spin-wave traps,” Sci. Rep. 5, 12824 (2015). dimensional magnonic crystal,” J. Appl. Phys. 120, 043901 (2016).
6
P. Frey, A. A. Nikitin, D. A. Bozhko, S. A. Bunyaev, G. N. Kakazei, 21
D. Petti, S. Tacchi, and E. Albisetti, “Review on magnonics with engineered
A. B. Ustinov, B. A. Kalinikos, F. Ciubotaru, A. V. Chumak, Q. Wang, spin textures,” J. Phys. D: Appl. Phys. 55, 293003 (2022).
V. S. Tiberkevich, B. Hillebrands, and A. A. Serga, “Reflection-less width- 22
Z. K. Wang, V. L. Zhang, H. S. Lim, S. C. Ng, M. H. Kuok, S. Jain, and
modulated magnonic crystal,” Commun. Phys. 3, 17 (2020). A. O. Adeyeye, “Nanostructured magnonic crystals with size-tunable bandgaps,”
7
E. V. Tartakovskaya, A. S. Laurenson, and V. V. Kruglyak, “Wannier-Stark ACS Nano 4, 643–648 (2010).
23
ladder spectrum of Bloch oscillations of magneto-dipole spin waves in graded H. Yu, G. Duerr, R. Huber, M. Bahr, T. Schwarze, F. Brandl, and D. Grundler,
1D magnonic crystals,” Low Temp. Phys. 46, 830–835 (2020). “Omnidirectional spin-wave nanograting coupler,” Nat. Commun. 4, 2702
8
Z. Chen and F. Ma, “Skyrmion based magnonic crystals,” J. Appl. Phys. 130, (2013).
24
090901 (2021). D. Zhang, Q. Yang, Y. Jiang, M. Wang, S. Du, I. I. Syvorotka, and H. Zhang,
9
R. Kryshtal and A. Medved, “Nonreciprocity of spin waves in magnonic crystals “Laser-induced damage of garnet films grown by LPE,” Opt. Mater. 91, 268–273
created by surface acoustic waves in structures with yttrium iron garnet,” (2019).
J. Magn. Magn. Mater. 395, 180–184 (2015). 25
S. Daimon, R. Iguchi, K. Uchida, and E. Saitoh, “Laser micro-processing as a
10
K. Di, S. X. Feng, S. N. Piramanayagam, V. L. Zhang, H. S. Lim, S. C. Ng, and tool for constructing insulator-based magnonic crystal,” J. Phys. D: Appl. Phys.
M. H. Kuok, “Enhancement of spin-wave nonreciprocity in magnonic crystals 48, 164014 (2015).
via synthetic antiferromagnetic coupling,” Sci. Rep. 5, 10153 (2015). 26
K. H. Chi, Y. Zhu, and C. S. Tsai, “Two-dimensional magnonic crystal with
11
M. Mohseni, V. I. Vasyuchka, V. S. L’vov, A. A. Serga, and B. Hillebrands, periodic thickness variation in YIG layer for magnetostatic volume wave propa-
“Classical analog of qubit logic based on a magnon Bose–Einstein condensate,” gation,” IEEE Trans. Magn. 49, 1000–1004 (2013).
Commun. Phys. 5, 196 (2022). 27
J. W. Kłos, M. L. Sokolovskyy, S. Mamica, and M. Krawczyk, “The impact of
12
T. Schwarze and D. Grundler, “Magnonic crystal wave guide with large spin- the lattice symmetry and the inclusion shape on the spectrum of 2D magnonic
wave propagation velocity in CoFeB,” Appl. Phys. Lett. 102, 222412 (2013). crystals,” J. Appl. Phys. 111, 123910 (2012).
13
S. Tacchi, G. Duerr, J. W. Klos, M. Madami, S. Neusser, G. Gubbiotti, 28
A. Manzin, G. Barrera, F. Celegato, M. Coïsson, and P. Tiberto, “Influence of
G. Carlotti, M. Krawczyk, and D. Grundler, “Forbidden band gaps in the spin- lattice defects on the ferromagnetic resonance behaviour of 2D magnonic crys-
wave spectrum of a two-dimensional bicomponent magnonic crystal,” Phys. Rev. tals,” Sci. Rep. 6, 22004 (2016).
29
Lett. 109, 137202 (2012). S. Orfanidis, Electromagnetic Waves and Antennas (Sophocles J. Orfanidis,
14
X.-Y. Jiang, Metamaterial (IntechOpen, Rijeka, 2012). 2016).
15
G. Monsivais and C. Ordóñez-Romero, “Band structure and Wannier-Stark 30
N. J. Whitehead, S. A. R. Horsley, T. G. Philbin, and V. V. Kruglyak, “A
ladders in the spin wave spectrum,” J. Magn. Magn. Mater. 466, 150–163 (2018). Luneburg lens for spin waves,” Appl. Phys. Lett. 113, 212404 (2018).
16
M. A. Morozova, O. V. Matveev, Y. P. Sharaevskii, S. A. Nikitov, and 31
N. J. Whitehead, S. A. R. Horsley, T. G. Philbin, and V. V. Kruglyak, “Graded
A. V. Sadovnikov, “Nonlinear signal processing with magnonic superlattice with index lenses for spin wave steering,” Phys. Rev. B 100, 094404 (2019).
32
two periods,” Appl. Phys. Lett. 120, 122407 (2022). O. Dzyapko, I. V. Borisenko, V. E. Demidov, W. Pernice, and
17
N. Grigoryeva, A. Ustinov, and B. Kalinikos, “Observation of spin-wave enve- S. O. Demokritov, “Reconfigurable heat-induced spin wave lenses,” Appl. Phys.
lope solitons in periodic magnetic film structures,” JETP Lett. 88, 31–35 (2008). Lett. 109, 232407 (2016).
18
A. V. Drozdovskii and B. A. Kalinikos, “Observation of self-modulation spin- 33
M. Vogel, R. Aßmann, P. Pirro, A. V. Chumak, B. Hillebrands, and G. von
wave instability in a one-dimensional magnonic crystal,” JETP Lett. 95, 357–361 Freymann, “Control of spin-wave propagation using magnetisation gradients,”
(2012). Sci. Rep. 8, 11099 (2018).
19
A. V. Chumak, A. A. Serga, S. Wolff, B. Hillebrands, and M. P. Kostylev, 34
K. G. Fripp and V. V. Kruglyak, “Spin-wave wells revisited: From wavelength
“Design and optimization of one-dimensional ferrite-film based magnonic crys- conversion and Möbius modes to magnon valleytronics,” Phys. Rev. B 103,
tals,” J. Appl. Phys. 105, 083906 (2009). 184403 (2021).