Plasmonic Modulators Can Break the Wireless Terahertz Barrier

Modern telecommunications infrastructure relies on a broad range of technologies. But ironically, some of these technologies can’t readily communicate with each other.

The electrical signals used for wireless communications, for example, can’t just be shoved into the fiber-optic infrastructure that forms the backbone of modern networks. Instead, they must be first converted to light (and then back again). This important task is performed by a network component called an electro-optic (EO) modulator.

“All information that you have is in the electrical world, but once it leaves your house, it goes into fiber. So, you need components that can encode from the electrical to the optical world signals at enormous speed. That’s where the modulator comes in,” says Juerg Leuthold, the head of the Department of Information Technology and Electrical Engineering at ETH Zurich, the Swiss Federal Institute of Technology of Zurich.

Telecommunications providers hope that next-generation 6G networks will deliver wireless speeds up to a terabit per second, and possibly beyond. However, these fast wireless networks still need to connect with wired fiber-optic infrastructure. That means electro-optic modulators need an upgrade—or else they risk becoming a bottleneck.

Plasmonic EO Modulator Breakthrough

Leuthold is coauthor of a paper recently published by researchers at ETH Zurich and Polariton Technologies in Switzerland that demonstrated a plasmonic EO modulator capable of frequencies up to 1.14 terahertz. It also provided 3-decibel EO bandwidth at a frequency of 997 gigahertz. Put more simply, the modulator can process signals up to nearly a terahertz before significant signal degradation occurs.

That’s a big leap from the modulators commonly in use today. Most are based on materials like lithium niobate (LiNbO₃), indium gallium arsenide (InGaAs) and, more recently, silicon. Modulators using these materials typically have a frequency response that degrades when frequencies reach 60 to 100 GHz. The plasmonic EO modulator achieved a roughly tenfold improvement.

As you might expect, a plasmonic EO modulator works a bit differently from its predecessors.

Conventional modulators often rely on the Pockels effect, which describes how an applied electric field can change the refractive index of a nonlinear crystal material. The changes in the refractive index alter the light that passes through the material, making it possible to write an electrical signal into an optical signal.

Plasmonic modulators still use the Pockels effect, but the light directed into the modulator is transformed. “We take the photons, a red photon, convert it into a plasmon, and the plasmon propagates along the surface of a metal,” explains Leuthold.

Plasmons are the quanta of electron oscillations in a metal, and they have useful properties. When coupled with electromagnetic fields, they form surface plasmons that can concentrate energy into volumes smaller than the wavelength of light. These plasmonic waves propagate across metal structures.

A plasmonic modulator takes advantage of this by cutting tiny slots just 100 nanometers wide into gold. The slots are filled with an organic electro-optic material, which can change the refractive index of the light. Within these slots, the optical signal (carried by plasmons) and electrical signal interact, writing the electrical signal into the optical signal.

Because the slots are so small, the electrical field is enhanced by up to 35,000 times. That allows a far stronger interaction between the electrical and optical signals.

Commercializing Plasmonic Modulators

The demonstration of a plasmonic EO modulator that achieves frequencies up to 1 THz is the latest in a decade-long string of plasmonic modulator innovations out of ETH Zurich.

ETH Zurich researchers, including Leuthold, published a paper on the use of plasmons for electrical to optical conversation in 2015 and, at the time, predicted it could allow frequencies up to 1 THz. They’ve now shown that possibility to be a reality.

Plasmonic modulators are being commercialized by Polariton. Spun out of ETH Zurich in 2019, Polariton was cofounded by three former Ph.D. students who contributed to prior research: Wolfgang Heni, Benedikt Baeuerle, and Claudia Hoessbacher.

Polariton currently offers silicon and plasmonic EO modulators capable of up to 145 GHz. Baeuerle says the company has “engineering samples available in small quantities” capable of up to 1 THz.

Modulators like this will be required if next-generation 6G telecommunications networks hope to live up to lofty promises.

While no standards have been set for 6G networks, they’re expected to use terahertz frequencies to deliver data rates that may soar beyond one terabit. Traditional EO modulators (which, as mentioned, top off around 100 GHz) would become a pinch point if these high-speed networks were put into practice.

The technology also has a place in AI data centers. Data centers built for AI typically have clusters of GPUs connected by an internal fiber-optic network. And, just like any other fiber-optic network, an electro-optical modulator is required to convert electrical signals to light (or back). Polariton produces both modulators and transceivers (which convert signals in both directions).

“Our electro-optic modulator is a solution for the next generation of transceivers for data centers and AI cluster that require high-speed and compact integration,” says Baeuerle. He notes that high-speed transceivers, including “next-generation” 3.2T (terabits per second) transceivers, will push electro-optical bandwidth to new heights.

Data rates this high might seem outlandish, and to be clear, 6G continues to face significant hurdles. Even so, advancements like plasmonic EO modulators and transceivers set the foundations required for faster, more reliable telecommunications.

“We are prepared for the next generation in the wireless world,” says Leuthold.