🔬 optics

Can Second-Order Nonlinearity in Metal-Dielectric Metamaterials Pave a Way Toward Elusive Photonic Time Crystals?

This paper proposes utilizing ultrafast second-order nonlinearities in low-loss metal-dielectric metamaterials to overcome the limitations of current transparent conductive oxides, demonstrating that high-intensity pumping can generate wide momentum bandgaps and net parametric gain to enable the realization of photonic time crystals in the optical range.

Original authors: Jacob B Khurgin

Published 2026-07-28

📖 4 min read☕ Coffee break read

CC BY 4.0

Original authors: Jacob B Khurgin

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine the world of light not just as a beam traveling through space, but as a song playing through time. Usually, when we think of mirrors or lenses, we are talking about how light bounces off things or bends as it moves through space. But what if the "mirror" itself changed its mind while the light was passing through it? This is the wild world of time-varying photonics. Instead of a static wall, imagine a wall that suddenly turns transparent, or a river that instantly changes its speed, right in the middle of a swimmer's path.

When light hits a medium that changes its properties incredibly fast—faster than a single wave of light can wiggle—it doesn't just reflect or refract; it can get "time-reflected" or "time-refracted." Think of it like a runner on a treadmill that suddenly speeds up or slows down; the runner's stride gets messed up, creating a new kind of energy. Scientists call a specific, extreme version of this a Photonic Time Crystal (PTC). It's a state where light gets trapped in a loop of time, growing stronger and stronger, almost like a snowball rolling down a hill that keeps getting bigger without ever stopping. The big question is: Can we actually build one? The answer has been "no" for a long time because the materials we've tried to use are either too slow to change fast enough, or they get so hot they melt before the magic happens.

Enter a new proposal from Jacob B. Khurgin at Johns Hopkins University. This paper suggests a clever workaround to build a Photonic Time Crystal using a "metamaterial"—a man-made sandwich of layers—instead of the usual materials that have failed so far. The author argues that the old way of trying to do this with transparent conductive oxides (like the clear coatings on your phone screen) is a dead end because those materials are too lossy and slow. Instead, the paper suggests using a stack of alternating layers: a semiconductor (a type of crystal) and silver (a metal).

Here is the magic trick: The author proposes using a "pump" laser beam that hits the side of this sandwich, while the "probe" light (the one we want to manipulate) travels through it. By carefully choosing the angle and the type of light, the pump can change the material's properties using a super-fast "second-order" effect, while the silver layers keep the heat down and the probe light from getting absorbed. The paper calculates that with pump powers around 100s of GW/cm² (gigawatts per square centimeter), this setup could open a wide "momentum bandgap" (a forbidden zone for light) of tens of percents. Even more exciting, the math suggests that at these power levels, the light wouldn't just bounce around; it would actually gain energy, achieving a net parametric gain of about 50% over a tiny 70 femtosecond window.

The paper is careful to point out that this is a theoretical roadmap, not a finished product. It explicitly rules out the current favorite materials (like ITO) because they rely on "third-order" effects that are either too weak or too slow, and they get too hot too fast. The author admits that building this structure is tricky—it would require stacking about 10 periods of layers, each 200 nm wide, with a 193 nm thick semiconductor layer and a 7 nm silver layer. However, the paper suggests this could be done using a "chiplet transfer" process, essentially stacking tiny slices of material like a deck of cards.

Crucially, the author estimates that this design is robust enough to survive the heat. A single 50-fs, 1-TW/cm² pulse would only raise the temperature of the silver layers by about 3 K, and it would take at least 1,000 pulses to raise the overall temperature by 10 K. This is a stark contrast to the old materials, which would likely melt or evaporate under similar conditions. The paper concludes that while we haven't built this yet, the path is clear: by separating the pump and probe light and using a metal-semiconductor sandwich, we might finally be able to catch a glimpse of a Photonic Time Crystal in the optical range, turning the elusive "time reflection" from a microwave curiosity into an optical reality.

Technical Summary: Second-Order Nonlinearity in Metal-Dielectric Metamaterials for Photonic Time Crystals

The Problem
The realization of Photonic Time Crystals (PTCs) in the optical range has been hindered by the inability to achieve strong, ultrafast modulation of a medium's permittivity while maintaining low optical loss. Current practical implementations rely heavily on Transparent Conductive Oxides (TCOs) like ITO and AZO. These materials suffer from two critical limitations:

Thermal Degradation and Loss: TCOs rely on third-order nonlinearity ( $\chi^{(3)}$ ) driven by free-carrier absorption. This process is inherently lossy, leading to rapid lattice heating and material damage at the high pump intensities required.
Speed vs. Strength Trade-off: The ultrafast component of $\chi^{(3)}$ is too weak to induce significant modulation, while the strong thermal component is too slow (tens to hundreds of femtoseconds) to support single-cycle or sub-cycle dynamics necessary for PTC formation.
ENZ Limitations: While Epsilon-Near-Zero (ENZ) materials offer large relative permittivity modulation, the "giant" nonlinearity in these materials is tied to strong absorption, creating a bottleneck where the pump energy required to drive the effect destroys the sample before a true PTC can be observed.

Methodology
The author proposes an alternative platform utilizing a metal-dielectric metamaterial to bypass the limitations of bulk TCOs. The core strategy involves:

Material System: An anisotropic metamaterial composed of alternating layers of a semiconductor (specifically (001)-oriented Al $_{0.75}$ Ga $_{0.25}$ As) and a metal (Silver).
Mechanism: Exploiting the material's second-order nonlinearity ( $\chi^{(2)}$ ) rather than $\chi^{(3)}$ . $\chi^{(2)}$ is inherently ultrafast and stronger, making it ideal for parametric processes.
Geometric Decoupling: A transverse pumping geometry is employed where the pump and probe waves propagate orthogonally.
- The pump is polarized normal to the layers (along the metal-semiconductor interface) to minimize absorption in the silver and avoid two-photon absorption in the semiconductor.
- The probe operates in the ENZ regime (near-zero effective permittivity) but is polarized parallel to the layers to ensure low loss and long interaction lengths.
Fabrication Approach: The paper outlines a feasible fabrication route using carrier-assisted sequential chiplet transfer to stack alternating 193 nm semiconductor and 7 nm silver layers, creating a 10-period structure with a total thickness of ~2 $\mu$ m.

Key Contributions and Results
The paper provides order-of-magnitude estimates and theoretical analysis demonstrating the viability of this approach:

Low Loss and Thermal Management: By separating the pump and probe polarizations, the pump absorption is minimized (absorption length ~1 mm), and the probe absorption is exceptionally low (absorption length ~2.14 $\mu$ m). This prevents the catastrophic thermal buildup seen in TCOs. Calculations suggest a single 1 TW/cm $^2$ pulse raises the silver layer temperature by only ~3 K, allowing for thousands of pulses before thermal equilibrium is reached.
Large Momentum Bandgap: With pump intensities of 100s of GW/cm $^2$ (specifically 1 TW/cm $^2$ in the primary estimate), the model predicts a permittivity modulation depth ( $\Delta\epsilon$ ) of ~0.3. This yields a fractional momentum bandgap of approximately 26% (or ~8% at 100 GW/cm $^2$ ), which is well within the range of experimental observability.
Net Parametric Gain: The analysis shows that the parametric gain exceeds the inherent losses. For a 70 fs interaction time (approx. 10 optical cycles), a net temporal gain of ~50% is achievable. This is a critical threshold, as TCOs typically fail to achieve net gain due to excessive loss.
Interaction Duration: The low loss allows the probe pulse to remain within the 2 $\mu$ m sample for ~74 fs, satisfying the requirement for multi-cycle interaction necessary to observe time-reflection and time-refraction phenomena.

Significance and Claims
The paper claims to offer a "robust pathway" to realizing Photonic Time Crystals in the optical range, a feat previously considered exceptionally difficult.

Paradigm Shift: It argues that the field should move away from the "impregnable obstacle" of TCO-based $\chi^{(3)}$ approaches and embrace $\chi^{(2)}$ in engineered metamaterials.
Feasibility: Unlike previous theoretical proposals that rely on "sheer willpower" regarding material properties, this work relies on established material parameters (Ag and AlGaAs) and realistic fabrication techniques (chiplet stacking).
Modest Outlook: The author maintains a modest tone, acknowledging that while the theoretical pathway is clear, experimental demonstration faces challenges. The work is presented as a necessary alternative to standard TCO approaches, aiming to guide the community toward a solution that balances ultrafast speed, high nonlinearity, and low thermal loss. The ultimate goal is to enable the observation of true time-refraction and time-reflection in the optical domain, phenomena currently restricted to the microwave regime.

More like this