Letters

https://doi.org/10.1038/s41567-019-0511-y

Chiral topological semimetal with multifold band

crossings and long Fermi arcs
Niels B. M. Schröter 1*, Ding Pei2, Maia G. Vergniory3,4, Yan Sun 5, Kaustuv Manna 5,
Fernando de Juan3,4,6, Jonas. A. Krieger 1,7,8, Vicky Süss5, Marcus Schmidt5, Pavel Dudin 9,
Barry Bradlyn 10, Timur K. Kim 9, Thorsten Schmitt1, Cephise Cacho9, Claudia Felser 5,
Vladimir N. Strocov1 and Yulin Chen2*

Topological semimetals in crystals with a chiral structure surface enclosing them in momentum space. This charge has integer
(which possess a handedness due to a lack of mirror and magnitude C = ±1, which is known as the Chern number and gives
inversion symmetries) are expected to display numerous the electronic wavefunction a handedness. This topological prop-
exotic physical phenomena, including fermionic excitations erty results in a plethora of exotic phenomena, such as Fermi-arc
with large topological charge1, long Fermi arc surface states2,3, surface states16, unconventional magnetoresistance12,17,18, non-local
unusual magnetotransport4 and lattice dynamics5, as well transport19 and many other nonlinear and anomalous responses20–25.
as a quantized response to circularly polarized light6. So far, Although relativistic Weyl fermions must obey Lorentz symmetry,
all experimentally confirmed topological semimetals exist in condensed matter systems are free from this constraint and instead
crystals that contain mirror operations, meaning that these need only be invariant under the space group of the crystal. This
properties do not appear. Here, we show that AlPt is a structur- freedom gives rise to a much richer variety of topological semimet-
ally chiral topological semimetal that hosts new four-fold and als, such as tilted Weyl semimetals26 and topological semimetals
six-fold fermions, which can be viewed as a higher spin gener- involving three, four or six band crossings protected by symme-
alization of Weyl fermions without equivalence in elementary try1,27,28. When such multifold crossings are realized in chiral crystal
particle physics. These multifold fermions are located at high structures, the effective Hamiltonian near the crossings is a higher
symmetry points and have Chern numbers larger than those spin generalization of the Weyl Hamiltonian H = k·S. Here, k is the
in Weyl semimetals, thus resulting in multiple Fermi arcs that crystal momentum and the matrices S are adiabatically connected
span the full diagonal of the surface Brillouin zone. By imaging to spin matrices for spin-1, spin-3/2 or the direct sum of two spin-1
these long Fermi arcs, we experimentally determine the mag- copies for the three-fold, four-fold and six-fold crossings, respec-
nitude and sign of their Chern number, allowing us to relate tively1. The bands of these multifold fermions feature higher Chern
their dispersion to the handedness of their host crystal. numbers (which also bestows their wavefunction with a handed-
An object that cannot be superimposed with its mirror image ness), and therefore host more Fermi arcs and chiral Landau levels
is said to be chiral, a concept first proposed by Lord Kelvin that (which are also qualitatively different1) compared to Weyl semimet-
has found widespread applications across the modern sciences, als, which enhances many of the phenomena associated with Weyl
from high-energy physics to biology. In condensed-matter physics physics. Spin-1 fermions, in particular, are markedly different from
(where the properties of crystalline materials are tightly constrained the triple band crossings recently discovered in tungsten carbide29
by spatial lattice symmetries), chiral crystals can only be found in and molybdenum phosphide30, which are not structurally chiral and
the 65 Sohncke space groups, which contain only orientation-pre- feature no well-defined Chern numbers.
serving operations and can therefore be assigned a handedness. The A non-zero Chern number does not require a chiral crystal struc-
structural chirality of these systems can endow them with fascinat- ture, but topological semimetals that also possess structural chiral-
ing properties, such as natural optical activity, negative refraction ity have many unique properties31 (see Fig. 1a for an overview), such
in metamaterials7, non-reciprocal effects such as magnetochiral as the gyrotropic magnetic effect4, a finite frequency version of the
birefringence of light or electronic magnetochiral anisotropy8,9, chi- chiral magnetic effect first predicted for the quark-gluon plasma32;
ral magnetic textures such as helices and skyrmions10 or unusual the signatures of the chiral anomaly that arise in phonon dynamics5
superconductivity11. In the recently discovered Weyl semimetals, or the magnetochiral anisotropy9; and most notably the quantized
not just the crystal structure itself, but also the electronic wave- circular photogalvanic effect6,33, the only experimental probe that
functions can exhibit a chirality at point-like two-band crossings can directly measure the Chern numbers of a topological semi-
of the quasiparticle dispersion, which are known as Weyl fermi- metal. However, all experimentally confirmed topological semimet-
ons12–15. These crossings are topologically protected because they als—of the Weyl type or any generalization—have been synthesized
carry a topological charge, a quantized Berry flux through any so far only in non-chiral crystal structures. Orientation reversing

1
Swiss Light Source, Paul Scherrer Institute, Villigen, Switzerland. 2Clarendon Laboratory, Department of Physics, University of Oxford, Oxford, UK.
3
Donostia International Physics Center, Donostia-San Sebastian, Spain. 4IKERBASQUE, Basque Foundation for Science, Bilbao, Spain. 5Max Planck Institute
for Chemical Physics of Solids, Dresden, Germany. 6Rudolph Peierls Centre for Theoretical Physics, University of Oxford, Department of Physics, Clarendon
Laboratory, Oxford, UK. 7Laboratory for Muon Spin Spectroscopy, Paul Scherrer Institute, Villigen, Switzerland. 8Laboratorium für Festkörperphysik, ETH
Zurich, Zurich, Switzerland. 9Diamond Light Source, Didcot, UK. 10Department of Physics and Institute for Condensed Matter Theory, University of Illinois
at Urbana-Champaign, Urbana, IL, USA. *e-mail: [email protected]; [email protected]

Nature Physics | VOL 15 | AUGUST 2019 | 759–765 | www.nature.com/naturephysics 759

Letters NATure PHySiCS

a b Non-chiral topological c d
Topological (110) (111) Four-fold
Chiral crystals semimetal b3
semimetals fermion
Mirror Pt C = +4
Optical activity Weyl fermions
Magnetochiral Unusual magneto- E Al
anisotropy resistance
Magnetic helices Nonlinear and ano- EC C– C+
and skyrmions malous Hall effects M
Γ Γ
Unusual k . S fermions Chern b1

Berr
supercon- numbers + k b2
Large Chern numbers + (001) X
ductivity largest Fermi arcs Fermi arcs

y flux
Chiral topological
Negative Unusual phonon Non-local semimetal
refraction dynamics transport
E Six-fold
Gyrotropic magnetic c fermion
effect EC+ C+ R C = –4
Quantized circular EC– C– a
photogalvanic effect b

Chiral topological
semimetals k

e f g h
1.5
X Γ
R X R R
1 R X R Γ
Pt 4f

3
Al 2p

0
0.5 X
M
Band
Γ 0 –0.5 crossing
Intensity (a.u.)

Γ
ky (Å–1)

E – EF (eV)
R
–0.5
20 30 40 50 60 70 80 90 –1
–76 –72 2θ (°) –1
1
Al 2s –1.5
–1.5
Pt 5p VB

.5
–1

.5

0
–2

–1

–0
0 1 –1
kx (Å–1) 0.5 –0.5
0 0.5 0
–0.5
–120 –100 –80 –60 –40 –20 0 –2 –1 0 1 2 –1 1 ky (Å–1)
Min Max kx (Å–1)
E – EF (eV) kx (Å–1)

Fig. 1 | Basic characteristics of chiral topological semimetals and AlPt. a, Overview of various exotic phenomena in chiral crystals, topological semimetals
and chiral topological semimetals. See also ref. 31. b, Energy positions of band crossings with opposite Chern number (C±) in non-chiral topological
semimetals (top) and chiral topological semimetals (bottom). c, AlPt crystal structure and possible cleavage planes. d, The cubic Brillouin zone of
AlPt, featuring an illustration of four-fold and six-fold fermions. e, Experimental core-level spectrum and powder diffraction pattern. Inset: green marks
indicate calculated Bragg peaks for AlPt. f, Experimental Fermi surface of the (110) cleavage surface, including the boundary of the bulk Brillouin zone,
taken at hv = 700 eV with right-circular polarization. g, Experimental Fermi surface of the (111) cleavage surface, taken with hv = 800 eV with left-circular
polarization. h, 3D intensity plot of our ARPES data measured with hv = 554 eV and linear-vertical polarization, showing a band crossing at the R point.

operations such as mirror symmetries in these materials require electronic structures, as well as soft X-ray energies (SX-ARPES,
the existence of Weyl nodes with opposite Chern numbers at the hv > 300 eV) to study their bulk band structures. To understand the
same energy (Fig. 1b), which force the aforementioned properties to experimental results, we also performed ab initio band structure
vanish. In a chiral topological semimetal (that is, a semimetal with calculations (see Methods for more details).
non-zero Chern number and broken mirror and inversion symme- AlPt crystallizes in the cubic chiral B20 structure of FeSi, which
try), nodes of opposite Chern number are generically separated in can be considered as a distorted rocksalt structure where the atomic
energy. The absence of a viable material to observe these proper- positions are displaced along the (111) direction with a lattice
ties has severely limited progress in the field up to now, despite an constant of a = 4.863. Because the material is not layered, it can
extensive theoretical search for them31. Promising material candi- be cleaved along different Miller planes, such as (001), (110) and
dates that could display these phenomena were recently predicted (111), as illustrated in Fig. 1c. Its cubic Brillouin zone is shown in
in a family of transition-metal silicides and platinum and palladium Fig. 1d, which also illustrates the four-fold and six-fold crossings
alluminides1–3, which crystallize in the chiral space group 198. In at the Γ and R points that were predicted in refs. 1 and 2, as well
this family of materials, a spin-3/2 fermion is realized near the as the Brillouin zone boundary for planes parallel to the (111) and
Fermi level at the Γ point (which corresponds to the elusive Rarita– (110) Miller planes. We confirmed its crystal structure and elemen-
Schwinger fermion predicted in particle physics), while a double tal composition via X-ray powder diffraction and core-level photo-
spin-1 fermion is realized at the R point. The Chern number count emission spectroscopy, respectively; the results are shown in Fig. 1e
predicts at least four chiral surface Fermi arcs that traverse the diag- (for more details about the samples, including information about the
onal of the surface Brillouin zone, thus connecting the projections presence of a small impurity phase, see Supplementary Materials).
of the Γ and R points, which would span the largest portion of the The experimental Fermi surfaces measured on the (110) and (111)
Brillouin zone of any topological material. In addition, the nodes surface are displayed in Fig. 1f,g, respectively, showing agreement
at Γ and R have a significant energy separation, which is a practi- with the expected rectangular and trigonal Brillouin zone bound-
cal advantage in realizing the physical phenomena discussed above. ary. Figure 1h also displays the experimental band dispersion near
Because sample surfaces of the silicides have proven difficult to pre- the R point, clearly showing the expected band crossing at R—the
pare by in situ crystal cleaving, we chose to study the material AlPt. first indication for the existence of six-fold fermions in AlPt.
In this work we conducted angle-resolved photoelectron spec- To further investigate these exotic new fermions, we performed a
troscopy (ARPES) measurements of AlPt crystals at low photon detailed comparison between the experimental bulk band structure
energies (VUV-ARPES, hv < 150 eV) to investigate their surface obtained with SX-ARPES and our ab initio calculations (Fig. 2). We

760 Nature Physics | VOL 15 | AUGUST 2019 | 759–765 | www.nature.com/naturephysics

NATure PHySiCS Letters
a b
X Г
R
R Г
X R X X R X X
0 (i) 0 (ii) Six-fold 0
–0.5 –0.5 fermion

–1 –1
E – EF (eV)

–1.5 –1.5
–2 –2 –0.3
–2.5 –2.5
–3 –3
–3.5 –3.5 –0.6

E – EF (eV)
–4 –4
–1 –0.5 0 0.5 1 –1 –0.5 0 0.5 1
ky (Å–1) ky (Å–1)
c 1 1
–0.9
(i) E = EF X
(ii)
0.5 0.5
–1.2
ky (Å–1)

0 0 R

–0.5 –0.5 –1
–1.5 –0.5
1 0
X 0.5 0 0.5
–1 –1 –0.5 –1 1 ky (Å–1)
–1 –0.5 0 0.5 1 –1 –0.5 0 0.5 1 kx (Å–1)
kx (Å–1) kx (Å–1)
d e M Г M M Г M
1 1
Four-fold
(i) E – EF = –0.77 eV (ii) Six-fold (i) (ii) fermion
fermion 0 0
0.5 0.5
–0.5 –0.5
E – EF (eV)
ky (Å–1)

0 0
–1 –1

–0.5 –0.5
–1.5 –1.5

–1 –1 –2 –2
–1 –0.5 0 0.5 1 –1 –0.5 0 0.5 1 1 1.5 2 2.5 1 1.5 2 2.5
kx (Å–1) kX (Å–1) ky (Å–1) ky (Å–1)

Fig. 2 | Bulk electronic structure of AlPt. a, Comparison between experimental (i) and calculated (ii) band dispersion along X–R–X, where the red
arrow indicates the six-fold band crossing, measured with hv = 554 eV. Note that all bands along this high symmetry direction are doubly degenerate.
b, Energy evolution of the constant-energy surfaces at the R point, taken with hv = 554 eV. Red dashed lines are a guide for the eye. c,d, Comparison
between the experimental and calculated Fermi- and constant-energy surface (which are shown for binding energy Eb = 0.77 eV for the experimental
data and Ec = 0.69 eV for the calculated data), measured with hv = 554 eV. The calculation shows additional hole pockets at the Fermi level that are
suppressed in the experiment due to matrix element effects in the photoemission process. e, Comparison between experimental and calculated band
dispersion along M–Γ–M, where the red arrow indicates the four-fold band crossing, taken with hv = 680 eV. All experimental data were measured with
linear-vertical polarization.

focus here primarily on data obtained from the (111) surface (fur- experimental Fermi surface in Fig. 2c(i) and constant-energy sur-
ther results on other surfaces are provided in the Supplementary face in Fig. 2d(i) are also in qualitative agreement with the calcula-
Materials). Figure 2a–d displays data obtained with a photon tions shown in Fig. 2c(ii) and 2d(ii), respectively, which confirms
energy of hv = 554 eV, which corresponds to a measurement plane the position of the six-fold fermion at R and reflects the symmetry
in the Brillouin zone that contains the R and X points, as shown expected from this measurement plane. Small quantitative devia-
by the orange dashed lines in Figs. 1d and 2c(ii), while Fig. 2e was tions (for example, the apparent absence of the predicted small
obtained with hv = 680 eV to access the Γ and M points (details hole pockets in the experimental data) are due to matrix element
about the determination of the measurement planes along kz are effects in the photoemission process not included in the calcula-
provided in the Supplementary Materials). The experimental dis- tions, which do not influence our conclusion about the existence of
persion along X–R–X shown in Fig. 2a(i) is in qualitative agreement the six-fold fermion at R. Figure 2e(i) shows the dispersion along
with our ab initio calculations shown in Fig. 2a(ii), which highlights the M–Γ–M plane, which is also in excellent agreement with the
that the experimentally observed nodal point consists of a six-fold calculation shown in Fig. 2e(ii), and indicates the expected four-
crossing of three doubly degenerate bands (the degeneracy is lifted fold fermion at Γ. Overall, the agreement between our SX-ARPES
along other momentum directions). The observed position of the data and our ab initio calculations clearly establishes the existence
nodal point at binding energy Eb ≈ 0.77 eV is close to the calculated of four-fold and six-fold fermions in AlPt.
value at binding energy Ec ≈ 0.69 eV. The experimental constant- To investigate the topological nature of these fermions, we com-
energy surfaces displayed in Fig. 2b show that this node exists as pare in Fig. 3 the experimental surface electronic structure of the
a band crossing along all in-plane momentum directions and that (001) cleavage plane, which is expected to host topological surface
the bands involved in the six-fold crossing disperse almost linearly Fermi arcs, with the respective experimental bulk electronic struc-
towards the Fermi level, leading to an electron pocket at R. The ture and ab initio calculation. Figure 3a shows the position of the

Nature Physics | VOL 15 | AUGUST 2019 | 759–765 | www.nature.com/naturephysics 761

Letters NATure PHySiCS

a b Projected
hv = 460 eV, bulk FS bulk gap Projected
R bulk
3.5
pockets

Max Γ
3
ky (Å–1)

kx kz
2.5
Min
ky

Bulk
gap
–1 –0.5 0 0.5 1
Bulk
kx (Å–1) pockets Bulk
pockets

c Bulk
Fermi arcs Surface Surface
hv = 145 eV, spin-up spin-down
bulk + surface FS
–1.5

–2

–2.5 Fermi arcs

Γ
ky (Å–1)

X1

–3

–3.5 R
Bulk
pockets

–4
–3.5 –3 –2.5 –2 –1.5 –1 –0.5 0 0.5 1
kx (Å–1)

d Fermi arcs e
0 –1.5 Max
E – EF (eV)

–0.5 –2
Min
(i)
ky (Å–1)

–1
–2.5 Fermi arcs
0
Surface
E – EF (eV)

–3
–0.5

(ii) Bulk (iv)

–1 –3.5
X1 R X1 –1 –0.5 0 0.5 1 Bulk
pockets
ky kx (Å–1)

Fig. 3 | Bulk electronic structure and surface Fermi arcs on the (001) cleavage plane. a, Bulk Fermi surface probed at hv = 460 eV in the kz = π/a plane.
b, Calculated 3D bulk Fermi surface of AlPt (hole pockets in magenta and electron pockets in cyan). The blue contours on top show the projection of the
bulk pockets to the (001) surface Brillouin zone. The red surface on the bottom approximately corresponds to the measurement plane at hv = 460 eV shown
in a. c, Experimental Fermi surface map measured with hv = 145 eV and linear-horizontal polarization, integrated over 100 meV at the Fermi level. Also
shown are ab initio slab calculations (blue and red lines), as well as the projected bulk calculations (blue contours). Both the experimental data and the
slab calculation show surface Fermi arcs threading through the projected bandgaps and spanning the full diagonal of the surface Brillouin zone (red dashed
lines are a guide for the eye). d, (i) Experimental band dispersion along the X 1−R 1 direction (raw data on the left and curvature plot on the right). (ii)
Corresponding ab initio slab calculation. e, Curvature plot of the experimental data shown in c; the red dashed line is a guide to the eye for the Fermi arc.

bulk Fermi surface pockets in the kz = π/a plane measured with bulk Fermi surface for kz = 0). These bulk pockets are well described
SX-ARPES, which suppresses contributions from surface states due by the bulk density functional theory (DFT) calculations shown
to the large probing depth (also see Supplementary Fig. 2b for the in Fig. 3b, where the measurement plane of Fig. 3a is indicated

762 Nature Physics | VOL 15 | AUGUST 2019 | 759–765 | www.nature.com/naturephysics

NATure PHySiCS Letters
a Right-handed
Al-helix along (111) C=+4 1 2
Cut 1 Cut 2
(ii) (iii) E (iv) E
R

C = +2

C = –2
(i) (111)
Left-moving
modes
Γ

Bulk gap
Right-moving
Pt

ky
0 0 modes
C = –4
Al

R C = +4
kx –π/a 0 π/a –π/a 0 π/a
ky ky

b C = –4
Left-handed 1 2 Cut 1 Cut 2
Al-helix along (111) (ii) R E (iv) E
(iii)

C = +2
(111) C = –2 Right-moving
(i) modes
Γ

Bulk gap
Left-moving
ky

0 0 modes
C = +4

R
C = –4
kx –π/a 0 π/a –π/a 0 π/a
ky ky

c d Cut 1 Right-moving e Cut 2

modes Left-moving
–1.5 (i) (i) modes
0 0
E – EF (eV)

–2
–0.5 –0.5
ky (Å–1)

–2.5
–1 –1
Right-moving Left-moving
modes modes
–3 (ii) (ii)
0 0
BZ boundary
BZ boundary
1

E – EF (eV)

–3.5
–1 –0.5 0 0.5 1 –0.5 –0.5
kx (Å–1)

Min Max
–1 –1
–3 –2.5 –2 –1.5 –3 –2.5 –2 –1.5
ky (Å–1) ky (Å–1)

Fig. 4 | Connection between the directionality of the surface Fermi arcs and the handedness of the bulk crystal. a,b, (i) Crystal structures of two
enantiomers (mirror images) of AlPt, which can be distinguished by the handedness of the helix formed by Al atoms along the (111) direction. (ii)
Corresponding configuration of the Chern number signs, Fermi arc dispersion and the flux of the Berry curvature (indicated by arrows and red/blue colour,
flowing from positive to negative Chern number). (iii,iv) Dispersion of the corresponding Fermi arcs along the line-cuts shown as red arrows in (ii).
c, Enlarged view of the experimental data shown in Fig. 3c, red arrows show position of the two line-cuts shown in d and e. d,e, Two line-cuts to the left and
to the right of the Γpoint, respectively: (i) raw data and (ii) curvature plot. Red dashed lines are a guide to the eye of the dispersion of the surface Fermi
arcs. Black dashed lines indicate the boundaries of the (001) surface Brillouin zone (BZ).

by the red surface (the experimental data also include contribu- Fig. 3a, we can identify large S-shaped Fermi arcs that are threading
tions from neighbouring planes due to kz broadening). We can see through these projected bulk gaps and are connecting the projected
from the projections of the bulk calculations to the (001) surface Γ and R points (highlighted by red dashed lines as guides to the
(shown as blue contours on top of Fig. 3b) that the gaps between eye), which are spanning the entire diagonal of the surface Brillouin
the bulk pockets shown in Fig. 3a remain gapped for all kz values zone and must originate from the surface. This conclusion is fur-
(that is, they are projected bulk gaps), meaning that no bulk bands ther corroborated by our ab initio slab calculations, which are also
are expected to appear in those gaps at any photon energy (because shown in Fig. 3c (showing surface spin-up and spin-down bands as
changing the photon energy would correspond to probing bands red and blue lines, respectively; bulk bands are shown in white), and
at different kz momenta). When comparing our surface-sensitive projected bulk band contours (solid blue lines), which match our
VUV-ARPES data shown in Fig. 3c to the bulk data presented in experimental data. To further enhance the contrast of the S-shaped

Nature Physics | VOL 15 | AUGUST 2019 | 759–765 | www.nature.com/naturephysics 763

Letters NATure PHySiCS

Fermi arcs, we present a curvature plot34 of the experimental Fermi 5. Rinkel, P., Lopes, P. L. S. & Garate, I. Signatures of the chiral anomaly in
surface in Fig. 3e, which also clearly shows the S-shaped Fermi phonon dynamics. Phys. Rev. Lett. 119, 107401 (2017).
6. de Juan, F., Grushin, A. G., Morimoto, T. & Moore, J. E. Quantized circular
arcs. We also present the experimental energy versus momen- photogalvanic effect in Weyl semimetals. Nat. Commun. 8, 15995 (2017).
tum dispersion of experimental data along the R −X 1 direction in 7. Pendry, J. B. A chiral route to negative refraction. Science 306,
Fig. 3d(i), which clearly shows the Fermi arc bands dispersing in the 1353–1355 (2004).
projected bulk bandgap; these are reproduced by the ab initio slab 8. Rikken, G. L. J. A., Fölling, J. & Wyder, P. Electrical magnetochiral anisotropy.
Phys. Rev. Lett. 87, 236602 (2001).
calculations in Fig. 3d(ii). Due to spin–orbit coupling, the surface
9. Morimoto, T. & Nagaosa, N. Chiral anomaly and giant magnetochiral
Fermi arcs must be spin-split, which is corroborated by the band anisotropy in noncentrosymmetric Weyl semimetals. Phys. Rev. Lett. 117,
splitting in our slab calculations. We therefore conclude that that 146603 (2016).
our surface-sensitive VUV-ARPES data, when contrasted with our 10. Mühlbauer, S. et al. Skyrmion lattice in a chiral magnet. Science 323,
bulk-sensitive SX-ARPES data and ab initio calculations, provide 915–919 (2009).
11. Carnicom, E. M. et al. TaRh2B2 and NbRh2B2: superconductors with a chiral
strong evidence for the existence of two surface Fermi arcs that are noncentrosymmetric crystal structure. Sci. Adv. 4, eaar7969 (2018).
approaching Γ and R , and which must be spin-split due to spin– 12. Nielsen, H. B. & Ninomiya, M. The Adler–Bell–Jackiw anomaly and Weyl
orbit coupling, which means that the new multifold fermions R and fermions in a crystal. Phys. Lett. B 130, 389–396 (1983).
Γ must have a Chern number with magnitude |C| = 4. 13. Xu, S.-Y. et al. Discovery of a Weyl fermion semimetal and topological Fermi
A further analysis of the Fermi arc dispersion can not only deter- arcs. Science 349, 613–617 (2015).
14. Lv, B. Q. et al. Experimental discovery of Weyl semimetal TaAs. Phys. Rev. X
mine the magnitude of the Chern numbers in AlPt, but also its sign, 5, 031013 (2015).
which is intimately related to the handedness of the crystal struc- 15. Yang, L. X. et al. Weyl semimetal phase in the non-centrosymmetric
ture. As we illustrate in Fig. 4a,b, the two enantiomers of AlPt can compound TaAs. Nat. Phys. 11, 728–732 (2015).
be distinguished by the handedness of the helix that is formed by 16. Wan, X., Turner, A. M., Vishwanath, A. & Savrasov, S. Y. Topological
their Al atoms along the (111) direction (which can be left-handed semimetal and Fermi-arc surface states in the electronic structure of
pyrochlore iridates. Phys. Rev. B 83, 205101 (2011).
or right-handed), as well as by the configuration of the signs of 17. Aji, V. Adler–Bell–Jackiw anomaly in Weyl semimetals: application to
their Chern number (either [Γ+/R−] or [Γ−/R+]). The fact that pyrochlore iridates. Phys. Rev. B 85, 241101 (2012).
the Chern number signs are inverted between the two enantiomers 18. Son, D. T. & Spivak, B. Z. Chiral anomaly and classical negative
can be understood by the fact that the integral of the Berry cur- magnetoresistance of Weyl metals. Phys. Rev. B 88, 104412 (2013).
vature acquires a minus sign upon a mirror operation because it 19. Parameswaran, S. A., Grover, T., Abanin, D. A., Pesin, D. A. & Vishwanath,
A. Probing the chiral anomaly with nonlocal transport in three-dimensional
involves the curl of the Berry connection. Which of these two sign topological semimetals. Phys. Rev. X 4, 031035 (2014).
configurations is realized for a particular enantiomer is not fixed by 20. Burkov, A. A. & Balents, L. Weyl semimetal in a topological insulator
symmetry, but can be calculated by ab initio calculations. Here we multilayer. Phys. Rev. Lett. 107, 127205 (2011).
find that the left-handed Al-helix corresponds to a positive Chern 21. Yang, K.-Y., Lu, Y.-M. & Ran, Y. Quantum hall effects in a Weyl
number at Γ, which means that the Berry curvature flows from Γ to semimetal: possible application in pyrochlore iridates. Phys. Rev. B 84,
075129 (2011).
R, which in turn defines the direction along which the Fermi arcs 22. Zyuzin, A. A. & Burkov, A. A. Topological response in Weyl semimetals and
disperse for line cuts between Γ and R : they become right-moving the chiral anomaly. Phys. Rev. B 86, 115133 (2012).
for a cut to the left of Γand left-moving for a cut to the right of 23. Grushin, A. G. Consequences of a condensed matter realization of
Γ. This directionality becomes reversed in the other enantiomer Lorentz-violating QED in Weyl semi-metals. Phys. Rev. D 86, 045001 (2012).
24. Goswami, P. & Tewari, S. Axionic field theory of (3 + 1)-dimensional Weyl
because the flow of Berry curvature is reversed. By examining our
semimetals. Phys. Rev. B 88, 245107 (2013).
experimental data (shown in Fig. 4c–e) along the two line cuts to the 25. Chan, C.-K., Lee, P. A., Burch, K. S., Han, J. H. & Ran, Y. When chiral
left and right of Γ that are passing through the projected bulk band- photons meet chiral fermions: photoinduced anomalous Hall effects in Weyl
gaps of AlPt (indicated by red arrows), we find that the Chern num- semimetals. Phys. Rev. Lett. 116, 026805 (2016).
ber sign configuration in the sample measured here was [Γ+/R−], 26. Soluyanov, A. A. et al. Type-II Weyl semimetals. Nature 527,
495–498 (2015).
and that the corresponding crystal structure involved a left-handed 27. Mañes, J. L. Existence of bulk chiral fermions and crystal symmetry. Phys.
Al helix. To the best of our knowledge, such a connection between Rev. B 85, 155118 (2012).
surface state dispersion and crystalline chirality has not been made 28. Wieder, B. J., Kim, Y., Rappe, A. M. & Kane, C. L. Double Dirac semimetals
before and can be used in the future to experimentally determine in three dimensions. Phys. Rev. Lett. 116, 186402 (2016).
the handedness of a chiral topological semimetal. 29. Ma, J.-Z. et al. Three-component fermions with surface Fermi arcs in
tungsten carbide. Nat. Phys. 14, 349–354 (2018).
Note added in proof: We recently became aware of other publi- 30. Lv, B. Q. et al. Observation of three-component fermions in the topological
cations that discuss chiral topological semimetals in the transition semimetal molybdenum phosphide. Nature 546, 627–631 (2017).
metal silicides35–37. 31. Chang, G. et al. Topological quantum properties of chiral crystals. Nat. Mater.
17, 978–985 (2018).
Online content 32. Fukushima, K., Kharzeev, D. E. & Warringa, H. J. Chiral magnetic effect.
Phys. Rev. D 78, 074033 (2008).
Any methods, additional references, Nature Research reporting 33. Flicker, F. et al. Chiral optical response of multifold fermions. Phys. Rev. B 98,
summaries, source data, statements of code and data availability and 155145 (2018).
associated accession codes are available at https://doi.org/10.1038/ 34. Zhang, P. et al. A precise method for visualizing dispersive features in image
s41567-019-0511-y. plots. Rev. Sci. Instrum. 82, 43712 (2011).
35. Sanchez, D. S. et al. Topological chiral crystals with helicoid-arc quantum
Received: 17 August 2018; Accepted: 22 March 2019; states. Nature 567, 500–505 (2019).
36. Takane, D. et al. Observation of chiral fermions with a large topological
Published online: 6 May 2019 charge and associated Fermi-arc surface states in CoSi. Phys. Rev. Lett. 122,
076402 (2019).
References 37. Rao, Z. et al. Observation of unconventional chiral fermions with long Fermi
1. Bradlyn, B. et al. Beyond Dirac and Weyl fermions: unconventional arcs in CoSi. Nature 567, 496–499 (2019).
quasiparticles in conventional crystals. Science 353, aaf5037 (2016).
2. Chang, G. et al. Unconventional chiral fermions and large topological Fermi Acknowledgements
arcs in RhSi. Phys. Rev. Lett. 119, 206401 (2017). The authors are grateful for excellent technical support from L. Nue and A. Pfister.
3. Tang, P., Zhou, Q. & Zhang, S.-C. Multiple types of topological fermions in The authors acknowledge Diamond Light Source for access to beamline I05 (proposal
transition metal silicides. Phys. Rev. Lett. 119, 206402 (2017). nos. SI19883 and SI21400) and the Paul Scherrer Institut, Villigen, Switzerland for
4. Zhong, S., Moore, J. E. & Souza, I. Gyrotropic magnetic effect and the provision of synchrotron radiation beam time at beamline ADRESS of the SLS.
magnetic moment on the Fermi surface. Phys. Rev. Lett. 116, 077201 (2016). N.B.M.S. acknowledges partial financial support from Microsoft. Y.L.C. acknowledges

764 Nature Physics | VOL 15 | AUGUST 2019 | 759–765 | www.nature.com/naturephysics

NATure PHySiCS Letters
support from the Engineering and Physical Sciences Research Council (grant no. EP/ wrote the manuscript with input and discussion from co-authors. V.N.S., C.F. and Y.C.
M020517/1). D.P. acknowledges support from the China Scholarship Council. F.J. supervised the research.
acknowledges funding from the European Union’s Horizon 2020 research and innovation
programme under Marie-Sklodowska Curie grant agreement no. 705968. J.A.K.
Competing interests
acknowledges the Swiss National Science Foundation (grant no. 200021_165910). Part
The authors declare no competing interests.
of the work of B.B. and M.G.V. was carried out at the Aspen Center for Physics, which is
supported by National Science Foundation grant PHY-1607611. M.G.V. was supported
by the IS2016-75862-P national project of the Spanish MINECO. K.M. and C.F. Additional information
acknowledge financial support from the ERC through grant no. 742068 ‘TOP-MAT’. Supplementary information is available for this paper at https://doi.org/10.1038/
s41567-019-0511-y.
Author contributions Reprints and permissions information is available at www.nature.com/reprints.
N.B.M.S. conducted the SX-ARPES experiments with the support of J.A.K. and V.N.S. Correspondence and requests for materials should be addressed to N.B.M.S. or Y.C.
and the VUV-ARPES experiments with the support of D.P. and P.D. The experimental
Journal peer review information: Nature Physics thanks David Carpentier and the other
data were analysed by N.B.M.S. and D.P. M.G.V. and Y.S. performed the VASP slab and
anonymous reviewer(s) for their contribution to the peer review of this work.
bulk ab initio calculations, and N.B.M.S. performed the Wien2k bulk calculations with
the support of D.P. F.J. and B.B. provided further theoretical support. K.M., V.S. and Publisher’s note: Springer Nature remains neutral with regard to jurisdictional claims in
M.S. grew the samples and K.M. performed the powder X-ray diffraction refinement. published maps and institutional affiliations.
P.D., T.K.K., T.S., C.C. and V.N.S. maintained the ARPES end stations. N.B.M.S. and F.J. © The Author(s), under exclusive licence to Springer Nature Limited 2019

Nature Physics | VOL 15 | AUGUST 2019 | 759–765 | www.nature.com/naturephysics 765

Letters NATure PHySiCS

Methods between top and bottom surface states, we considered a slab of 10 unit cells and
Sample growth. The polycrystalline sample of AlPt was first prepared by arc- 1 nm vacuum thickness. For the energy cuts, we used a 100 × 100 grid of k-points.
melting a stoichiometric amount of highly pure Al and Pt metals. The ingot was The spin resolved plots were done using PyProcar code46,47.
then crushed into fine powder and the final sample was prepared by one of two The Wien2k calculations employed a full-potential linearized augmented
methods. Before the final stage of sample synthesis, the compound’s melting profile plane-wave and local orbitals basis, as well as the PBE prescription of the exchange
was determined using differential scanning calorimetry. In the first technique, correlation term. The plane-wave cutoff parameter RMTKMAX was set to 7 and
crushed AlPt powder was placed in an aluminium tube and then sealed inside the irreducible Brillouin zone was sampled by 97,336 k-points. Spin–orbit coupling
a tantalum tube. The entire assembly was homogeneously heated in an argon was included via a second variational procedure. The results from the Wien2k
atmosphere up to 1,600 °C above the melting point of the compound and then calculations were exported with XCrySDen software48 into the bxsf file format to
cooled to 400 °C at a rate of 0.5 °C min−1. Finally, the sample was annealed at plot the 3D Fermi surface with MATLAB.
1,000 °C for 6 days followed by slow cooling at 1 °C min−1 to room temperature. In The crystal structure and Miller planes plotted in Fig. 1c were generated
the second method, we followed the Bridgman–Stockbarger technique to prepare with VESTA49.
the AlPt sample. The crushed polycrystalline powder was placed in a custom-
designed sharp-edged aluminium tube, which was sealed in a tantalum tube. The Data availability
entire assembly was then heated to 1,600 °C and then slowly cooled to 1,000 °C. The data that support the plots within this paper and other findings of this study
Compositional analysis, obtained with energy-dispersive X-ray spectroscopy, are available from the corresponding authors on reasonable request.
shows stoichiometric Al and Pt in a 1:1 composition in all samples. The samples
also show a similar pattern under laboratory X-ray powder diffraction analysis. We
used the AlPt samples prepared by both techniques for ARPES measurements. References
38. Strocov, V. N. et al. Soft‐X‐ray ARPES facility at the ADRESS beamline of the
ARPES. SX-ARPES measurements were performed at the SX-ARPES endstation38 SLS: concepts, technical realisation and scientific applications. J. Synchrotron
of the ADRESS beamline39 at the Swiss Light Source, Switzerland, with a SPECS Radiat. 21, 32–44 (2014).
analyser with an angular resolution of 0.07°. The photon energy varied from 350 eV 39. Strocov, V. N. et al. High-resolution soft X-ray beamline ADRESS at the Swiss
to 1,000 eV and the combined energy resolution ranged between 50 meV and Light Source for resonant inelastic X-ray scattering and angle-resolved
150 meV. The temperature during sample cleaving and measurements was ~20 K photoelectron spectroscopies. J. Synchrotron Radiat. 17, 631–643 (2010).
and the pressure was lower than 1 × 10−10 mbar. The increase of the photoelectron 40. Strocov, V. N. Three-dimensional electron realm in VSe2 by soft-X-ray
mean free path in the soft-X-ray energy range results (by the Heisenberg photoelectron spectroscopy: origin of charge-density waves. Phys. Rev. Lett.
uncertainty principle) in a higher kz resolution of the ARPES experiment compared 109, 086401 (2012).
to measurements at lower photon energies40, which was critical to measure the new 41. Hoesch, M. et al. A facility for the analysis of the electronic structures of
fermions in the bulk band structure of AlPt. solids and their surfaces by synchrotron radiation photoelectron spectroscopy.
VUV-ARPES measurements were performed at the high-resolution- and Rev. Sci. Instrum. 88, 013106 (2017).
nano-ARPES branch line of beamline I05 at the Diamond Light Source, UK41. 42. Kresse, G. & Furthmueller, J. Efficiency of ab-initio total energy calculations
Measurements at the high-resolution branch were performed with a Scienta R4000 for metals and semiconductors using a plane-wave basis set. Comput. Mater.
analyser, and a photon energy range between 130 eV and 160 eV, at a temperature Sci. 6, 15–50 (1996).
below 20 K. Measurements at the nano-ARPES branch were performed with a 43. Kresse, G. & Furthmueller, J. Efficient iterative schemes for ab initio
DA30 analyser at a photon energy of 82 eV, a sample temperature below 25 K, total-energy calculations using a plane-wave basis set. Phys. Rev. B 54,
and a beam spot size below 1 µm, which allowed us to measure the Fermi surface 11169 (1996).
map shown in Fig. 3c, which originates from a single domain with a diameter of 44. Blaha, P., Schwarz, K., Madsen, G. K. H., Kvasnicka, D. & Luitz, J. WIEN2k,
~15 µm. Measurements in the VUV-ARPES regime are more surface-sensitive than An Augmented Plane Wave+Local Orbitals Program for Calculating Crystal
SX-ARPES and therefore most suitable to image the Fermi arcs in AlPt. Properties (Karlheinz Schwarz Techn. Universität Wien, 2018).
45. Perdew, J. P., Burke, K. & Ernzerhof, M. Ceramic expansion by water layers
Ab initio calculations. We employed DFT, as implemented in the Vienna Ab Initio on magnesium oxide: ab initio study. Phys. Rev. Lett. 77, 3865 (1996).
Simulation Package (VASP)42,43 for the bulk calculations shown in Fig. 2 and the 46. Romero, A. H. & Munoz, F. Pyprocar Code https://github.com/romerogroup/
slab calculations shown in Fig. 3b–d, as well as Wien2k44 for the bulk calculations pyprocar (2015).
shown in Fig. 3a,b and Supplementary Fig. 2. 47. Singh, S., Garcia-Castro, A. C., Valencia-Jaime, I., Muñoz, F. & Romero, A. H.
For the VASP calculations, the exchange correlation term is described Prediction and control of spin polarization in a Weyl semimetallic phase of
according to the Perdew–Burke–Ernzerhof (PBE) prescription together with BiSb. Phys. Rev. B 94, 161116 (2016).
projected augmented-wave pseudopotentials45. For the autoconsistent calculations, 48. Kokalj, A. XCrySDen—a new program for displaying crystalline structures
we used a 7 × 7 × 7 k-points mesh for the bulk and 4 × 4 × 1 for the slab calculations. and electron densities. J. Mol. Graphics Model. 17, 176–179 (1999).
The kinetic energy cutoff was set to 400 eV. We calculated the surface states by 49. Momma, K. & Izumi, F. VESTA 3 for three-dimensional visualization of
using a slab geometry along the (001) direction. To achieve a negligible interaction crystal, volumetric and morphology data. J. Appl. Crystallogr. 44,
between the surface states from both sides of the slab and reduce the overlap 1272–1276 (2011).

Nature Physics | www.nature.com/naturephysics