Lightwave Logic Inc (LWLG)

[prototype_101]

Christian Reimer from HyperLight quote >> I wanted to kind of take back what we heard from the panelists and here at this show, and kind of see a lot of bottlenecks with this massive AI scaling, that we're going in all areas, from scale up, we're going to go from copper to optics…scale out, we're going from 200G to 400G, and scale across the telecom, we're going from 800 to 1.6G, so there's a big demand for speed and density.

But there's an even bigger demand for power, right? We heard that these data centers are ginormous in size in terms of gigawatt scale, and that the power demand doubles, more than doubles every year, which is faster than Moore's Law.

And a demand of, like, a staggering around a ***BILLION*** 1.6T equivalent ports will be needed by 2028, so every single picojoule per bit counts.

https://www.reddit.com/r/LWLG/comments/1rzh9ek/ofc_2026_tfln_photonics_at_the_inflection_point/

Ok, so let's discuss the pertinent FACTS

1) On Density LWLG can place 120 modulators in the area that TFLN can shoehorn in only 8 modulators

2) On Power LWLG will yield at least TWICE as the power savings, perhaps even greater System-wide

3) On Demand, "a staggering around a billion 1.6T equivalent ports will be needed by 2028" <

LWLG Polymer (LWLG/ GLOBAL FOUNDRIES/ TOWER)
Spin-on Liquid - (Simple)
300mm (12-inch) standard.
Standard CMOS etching.
Yield Potential - Higher (Standard semiconductor flow).

TFLN (HyperLight/UMC)
Material Application - Bonding & Slicing (Complex)
Wafer Size - 200mm (8-inch) is current limit.
Etching - Difficult; specialized Argon tools.
Yield Potential - Lower (due to crystal fragility)

Many ISSUES with TFLN in the CMOS Foundries!!! here>>>

The TFLN Foundry Process: Step-by-Step
Processing TFLN is significantly more complex than standard silicon because Lithium Niobate is a brittle crystal, not a liquid or a gas-grown layer.

Step A: Ion Slicing (The "Smart Cut" Process)
You cannot "grow" a thin film of Lithium Niobate on silicon. Instead, a bulk crystal of Lithium Niobate is bombarded with high-energy ions (usually Helium or Hydrogen). These ions create a "weakness plane" at a specific depth (e.g., 600 nanometers) within the crystal.

Step B: Wafer Bonding
The "sliced" bulk crystal is then flipped and bonded to a handle wafer (usually Silicon or Silicon Dioxide). This is a critical "fusion bonding" step that must be performed in a vacuum with zero dust particles; even a single microscopic speck will cause the thin film to crack or peel.

Step C: Thermal Exfoliation
The bonded "sandwich" is heated. The heat causes the ions trapped at the weakness plane to expand, "snapping" the thin film off the bulk crystal. What remains on the silicon wafer is a perfectly smooth, ultra-thin layer of Lithium Niobate (TFLN).

Step D: Specialized Dry Etching (The Hardest Part)
Lithium Niobate is notoriously difficult to etch. Standard silicon chemicals don't work. Foundries use Argon Ion Milling or Reactive Ion Etching (RIE) to "carve" the waveguides into the TFLN.

The Challenge: If the walls of the waveguide aren't perfectly smooth (atomic-level smoothness), the 1.6T signal will "leak" light, causing high insertion loss.

Step E: Metalization & Cladding
Once the waveguides are carved, the foundry adds a layer of cladding (usually SiO2) and then deposits metal electrodes (Gold or Copper). Because TFLN works on the Pockels Effect, these electrodes must be placed within microns of the waveguide to allow the electrical signal to change the refractive index of the crystal instantly, creating the 1.6T optical pulse.