confinement consistent with topological insulation. This constituted a
remarkable demonstration of a Floquet
TI in a physical system as well as an
optical topological insulator.
The proof-of-concept experiment
has raised many more questions: How
much can we perturb these systems
before we break topological insulation?
Do solitons exist in the TI gap? Are
delocalized states possible in a disordered photonic TI? How will entangled
photons behave? These are questions
that are easier to answer or only possible
to answer in photonic systems. We can
see many applications and fundamental
phenomena for this strange robustness
Topological insulators (TIs) insulate in their bulk and conduct electricity on their
surfaces. In 2-D TIs, the edge current has
scatter-free propagation, making it extremely
robust. Perhaps the most intriguing feature of
this new class of solids is that their physical
properties are related to topological quantities
rather than geometric ones, meaning they are
not affected by small perturbations. Besides
their impact on fundamental science, TIs may
lead to sophisticated applications in quantum
devices and spintronics.
Earlier this year, we formulated a proof-of-concept experiment demonstrating topological
robustness in optical physics. 1 The concept
of topological protection of electromagnetic
radiation was put forward by Haldane, and
experimentally demonstrated by Soljacic’s
group. 3 However, their approach relied on the
strong gyromagnetic response only achievable in the microwave regime. 2 We needed a
new methodology in order to scale down and
observe the topological protection of light. That
is why we used Floquet TIs. Pioneered by two
condensed matter physics groups, Floquet TIs
temporally modulate a solid in order to give a
preferred direction in time—i.e., to break time-reversal symmetry or Lorentz reciprocity. 3, 4
Fan’s group proposed similar concepts in order
to make optical isolators. 5
We used waveguides in a honeycomb
lattice. The propagation dynamics therein can
be described by a paraxial Schrödinger equation, where the spatial coordinate, z, takes the
place of time. By making the waveguides helical instead of straight, we effectively broke
z-reciprocity and allowed for scatter-free edge
states, and thus, we could directly observe
light transport around corners without
backscattering. We also observed light travel
around strong defects without disruption.
Furthermore, our experiments showed edge
(a) Observation of photonic topological protection in a honeycomb array of waveguides with a 15 μm neighbor distance and 8 μm helix radius. (b) A missing waveguide at the edge acts as a defect; yellow indicates the waveguide with injected
light. (c) Injected light moves clockwise and avoids the defect. Backscattering is
suppressed due to topological protection.
Mikael C. Rechtsman
Yaakov Lumer and
Technion, Haifa, Israel
Julia M. Zeuner and
Friedrich Schiller University
of Jena, Germany
1. M.C. Rechtsman et al.
Nature 496, 196 (2013).
2. Z. Wang et al. Nature
461, 772 (2009).
3. T. Kitagawa et al. Phys.
Rev. B 82, 235114 (2010).
4. N.H. Lindner et al. Nat.
Phys. 7, 490 (2011).
5. K. Fang et al. Nat. Pho-
tonics 6, 782 (2012).