INSIGHTS | PERSPECTIVES
GRAPHIC: A. MASTIN/SCIENCEscience.org SCIENCEessentially flipping a bit ( 2 ). To fully
realize this optical control of mag-
netism, scientists need to unravel
the fundamental physical processes
governing the transfer of energy
from the laser pulse to electrons
in the material. Recently, research-
ers have found ways to manipulate
magnetic orders inside materials
indirectly. Instead of directly excit-
ing magnons in a material, these
methods couple to magnetic order
by manipulating the crystal struc-
ture of the material with light ( 3 ).
The magnetic order of a ma-
terial depends on the geometry
and symmetry of its crystal lat-
tice, which affects the distances
between the atoms and thus the
quantum-mechanical exchange in-
teractions that are responsible for
magnetic ordering. Therefore, a
natural approach is to look at col-
lective vibrations of atoms, because
they are directly linked to changes
in the crystal structure and subse-
quently to the desired modulations
of the exchange interactions. These
crystal lattice vibrations follow de-
fined patterns of atomic motion,
and their quantized quasiparticles
are called “phonons.” Just like the photons
from a laser, phonons can interact with
magnetic order in a way similar to their lu-
minous siblings.
The spin-lattice interaction makes it pos-
sible to use phonons as the intermediary
between the laser pulse and the magnetic
order. However, efficiently transferring en-
ergy from photons to phonons has been a
long-standing challenge, because phonon
frequencies often lie in an electromagnetic
spectrum known as the “terahertz gap”
between 0.3 and 30 THz, where there is a
lack of practical technologies for generat-
ing and detecting electromagnetic waves.
This has been overcome in recent years,
and the transfer of energy from phonons to
magnons has been achieved experimentally
and theoretically ( 4 , 5 ). As powerful table-
top terahertz sources are becoming more
widely accessible, more demonstrations of
optically activated spin-lattice interactions
have been reported, such as those that show
the reorientations of magnetic domains or
the generation of magnetization in antifer-
romagnets (6–8). In these studies, the trans-
fer of energy is directed from structural to
magnetic order, but there is no fundamen-
tal reason why this could not go the other
way round as well. Because photon-magnon
coupling tends to be weaker than photon-
phonon coupling, a photon-to-magnon-
to-phonon system would require a more
efficient coupling mechanism between the
different quasiparticles.
Mashkovich et al. demonstrate an effi-
cient coupling between magnons and pho-
nons in the prototype antiferromagnetic
compound cobalt fluoride (CoF 2 ). They iden-
tify a magnon that couples its spin preces-
sion to the magnetic field of the light from
a laser pulse, which enables one to drive the
precession resonantly and with a large am-
plitude. This observation was linked to co-
herent structural oscillations in the crystal
corresponding to a phonon mode, which in
turn is coupled to the electron orbits and
the magnetization of the bulk material ( 6 ,
9 ). However, despite the precessing and
oscillating of the magnon and phonon at
the same time, there are still multiple pos-
sible channels through which the phonon
could have received its energy. How can
one be sure that the phonon is not being
driven by the light directly? To answer this,
Mashkovich et al. performed temperature-
dependence measurements of the excitation
mechanism. They found that the oscillation
of the phonon vanishes at the same temper-
ature at which the material loses its mag-
netic order, which strongly suggests that
the phonon receives most of its energy from
the magnetic system, through a previously
unknown energy-transfer mechanism and
with unprecedented efficiency.
From a theoretical perspective, we notethat the coupling of light, ma-
gnons, and phonons demonstrated
by Mashkovich et al. is reminiscent
of that recently predicted in the
infrared resonant Raman effect
for a photon-to-phonon-to-phonon
system ( 10 ), in which the light’s
electric field and the phonon’s elec-
tric dipole moment take the place
of the light’s magnetic field and
the magnon’s magnetic moment.
Reciprocally, the coupling demon-
strated by Mashkovich et al. could
be used to generate a magnetic
infrared resonant Raman effect,
which adds yet another route to
controlling magnetism through the
crystal lattice (see the figure).
Research on phonon-driven mag-
netic systems, also dubbed phono-
magnetism or magnetophononics,
has seen a rapid increase in recent
years, and this trend will likely ac-
celerate as more laboratories with
strong terahertz sources become
operational. Theoretical and com-
putational work plays a pivotal
role in furthering the fundamental
understanding and description of
these emerging phenomena, not
only reactively but also proactively
in predicting new physical mechanisms and
materials in which to find them (11–15).
A central challenge for the research field
is finding compounds with strong spin-
lattice couplings that can help achieve
terahertz-speed magnetic control. Here,
first-principles calculations combined with
high-throughput methods and artificial in-
telligence–driven techniques may be the
solution, because these methods can scan
over a vast range of possible candidate
compounds. The study of Mashkovich et
al. has set a precedent for where to start
the search. jREFERENCES AND NOTES- E. A. Mashkovich et al., Science 374 , 1608 (2021).
- S. Schlauderer et al., Nature 569 , 383 (2019).
- A. S. Disa, T. F. Nova, A. Cavalleri, Nat. Phys. 17 , 1087
(2021). - T. F. Nova et al., Nat. Phys. 13 , 132 (2017).
- D. M. Juraschek, M. Fechner, A. V. Balatsky, N. A. Spaldin,
Phys. Rev. Mater. 1 , 014401 (2017). - A. S. Disa et al., Nat. Phys. 16 , 937 (2020).
- D. Afanasiev et al., Nat. Mater. 20 , 607 (2021).
- A. Stupakiewicz et al., Nat. Phys. 17 , 489 (2021).
- P. G. Radaelli, Phys. Rev. B 97 , 085145 (2018).
- G. Khalsa, N. Benedek, J. Moses, Phys. Rev. X 11 , 021067
(2021). - M. Gu, J. M. Rondinelli, Phys. Rev. B 98 , 024102 (2018).
- M. Fechner et al., Phys. Rev. Mater. 2 , 064401 (2018).
- M. Rodriguez-Vega et al., Phys. Rev. B 102 , 081117(R)
(2020). - D. M. Juraschek, P. Narang, N. A. Spaldin, Phys. Rev. Res.
2 , 043035 (2020). - D. M. Juraschek, D. S. Wang, P. Narang, Phys. Rev. B 103 ,
094407 (2021).
10.1126/science.abm0085
+
_- _
Infrared-active
phononsMagnonsRaman-active
phononsInfrared resonant
Raman effectMagnetic
infrared resonant
Raman effectC o F Electric field
Magnetic fieldTwo pathways to the Raman effect
In Raman scattering, the electric field of the incident light is coupled
to infrared-active phonons, which convert the energy into lattice
vibrations. In the magnetic analog of the Raman effect, the magnetic
field of the incident light is coupled to magnons in the material and
generates lattice vibrations in a similar fashion.1556 24 DECEMBER 2021 • VOL 374 ISSUE 6575