🔬 applied physics

Demonstration of an Integrated Planar Guided-wave Terahertz Synthesized Filter

This paper demonstrates the successful application of microwave filter synthesis methods to design and experimentally validate integrated planar low-pass filters with a 0.8 THz cutoff frequency, achieving transmission characteristics that align with theoretical predictions.

Original authors: Ali Dehghanian, Mohsen Haghighat, Thomas Darcie, Levi Smith

Published 2026-03-27

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Ali Dehghanian, Mohsen Haghighat, Thomas Darcie, Levi Smith

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 you are trying to send a secret message using a very high-pitched whistle (Terahertz waves). This whistle is so high-pitched that it sits in a "gap" between the radio waves we use for Wi-Fi and the light we use for lasers. It's a mysterious, hard-to-reach place in the world of physics.

For a long time, scientists could build filters for radio waves (like in your phone) and light (like in fiber optics), but building a filter for these "Terahertz whistles" was like trying to build a house in a hurricane. It was too hard to measure the results, so most people just guessed how the filters worked instead of designing them precisely.

This paper is like a team of engineers saying, "We're going to stop guessing. We're going to build a custom filter for these high-pitched whistles, just like a master chef follows a recipe to bake the perfect cake."

Here is the breakdown of what they did, using some everyday analogies:

1. The Problem: The "Wild West" of High Frequencies

Think of the Terahertz range as a wild, uncharted territory.

The Old Way: Scientists usually built a weird-looking structure (like a maze of metal), then ran a test to see what happened. It was like throwing a dart at a board and hoping it hits the bullseye. They could analyze the result, but they couldn't design it to hit a specific target.
The Challenge: Measuring these waves is incredibly difficult. The equipment is expensive, the signals get lost easily, and the waves behave strangely.

2. The Solution: The "Recipe" (Synthesis)

The authors decided to use a method called Filter Synthesis.

The Analogy: Imagine you want a cake that tastes exactly like chocolate with a specific level of sweetness. Instead of just mixing random ingredients and hoping it tastes good, you use a precise recipe (math) to calculate exactly how much flour, sugar, and cocoa you need to get that specific flavor.
In the Paper: They took the "recipes" (mathematical formulas) used for radio waves and applied them to Terahertz waves. They wanted to build a Low-Pass Filter.
- What is a Low-Pass Filter? Think of it as a bouncer at a club. It lets the "slow" dancers (low-frequency signals) into the VIP section, but kicks out the "fast" dancers (high-frequency noise) before they can cause a mess.

3. The Design: The "Stepped Staircase"

To build this filter, they didn't use a solid block of metal. They used a Coplanar Stripline (CPS), which is basically a tiny, flat highway for electricity.

The Analogy: Imagine a road that changes its width. Some parts are wide and smooth (low impedance), and some parts are narrow and bumpy (high impedance).
The Magic: By alternating these wide and narrow sections, they created a "staircase" effect. When the electrical signal travels up this staircase, the high-pitched "fast" signals get confused and bounce back, while the "slow" signals glide right through.
The Specifics: They designed three different versions of this staircase (3 steps, 4 steps, and 5 steps). They chose a "Bessel" design, which is like a smooth, gentle ramp. This ensures that the message doesn't get distorted or "ring" like a bell when it passes through.

4. The Experiment: The "Flashlight and Mirror" Trick

How do you test a filter for waves that are too fast for normal electronics to catch?

The Setup: They used a special system called a "System-on-Chip."
- The Transmitter: A tiny switch that acts like a strobe light, flashing electricity incredibly fast (trillions of times a second).
- The Receiver: Another switch that acts like a camera shutter.
The Trick: They shine a laser pulse to trigger the transmitter. The electrical signal zooms through their filter. Then, the receiver catches it. But here's the cool part: they use a mechanical delay line (a mirror on a track) to change when the receiver looks. By moving the mirror back and forth, they can "reconstruct" the movie of the signal passing through the filter, frame by frame.

5. The Results: The "Perfect Match"

When they looked at the data:

Time Domain (The Movie): The signal came out looking almost exactly the same as when it went in. No distortion. It was like sending a clear voice through a tunnel and hearing it perfectly on the other side.
Frequency Domain (The Graph): The filter did exactly what the math predicted. It let the slow signals through and blocked the fast ones right at the "cut-off" point (0.8 Terahertz), just like a bouncer checking IDs at the door.

Why This Matters

This paper is a big deal because it's the first time someone has successfully used a precise "recipe" to build a Terahertz filter.

Before: It was like trying to build a car engine by guessing which parts fit together.
Now: We have a blueprint. We can now design these filters to do specific jobs, like cleaning up signals for future 6G internet, medical imaging, or security scanners.

In a nutshell: The authors proved that you can take the boring, reliable math used for radio waves and use it to build precise, custom tools for the mysterious world of Terahertz waves, opening the door for faster and clearer communication in the future.

1. Problem Statement

The THz Gap Challenge: While filter design is well-established for electrical (microwave) and optical frequencies, the Terahertz (THz) gap (0.1–10 THz) lacks validated experimental demonstrations of filter synthesis.
Analysis vs. Synthesis: Current literature in the THz regime primarily focuses on filter analysis (designing periodic structures or metastructures and characterizing their response post-fabrication) rather than filter synthesis (starting with a specific target transfer function, such as Chebyshev or Bessel, and calculating the geometry to achieve it).
Characterization Limitations: Validating THz devices is difficult due to:
- Electronic limitations: Vector Network Analyzers (VNAs) typically cap at 100 GHz; extension modules exist but are costly, dispersive, and band-limited.
- Optical limitations: Free-space THz optics suffer from field coupling errors and are difficult to manage with specific feedline geometries.
- Integration: Existing THz bulk optical components do not integrate well with miniaturized planar guided-wave transmission lines (TLs) like Coplanar Striplines (CPS) or Coplanar Waveguides (CPW).

2. Methodology

The authors applied microwave filter synthesis techniques to the THz domain using a System-on-Chip (TSoC) platform.

Platform & Device:
- Substrate: Thin Silicon Nitride ( $Si_3N_4$ , $H=1\mu m$ , $\epsilon_r=7.6$ ).
- Transmission Line: Coplanar Stripline (CPS) integrated with photoconductive switches (PCS) for DC biasing and signal generation/detection.
- Filter Topology: Stepped-Impedance Low-Pass Filters (LPF) designed with a Bessel response. Bessel was chosen for its linear phase response, short physical length, and minimal pulse distortion.
Design Process:
- Target Specifications: Cut-off frequency ( $f_c$ ) = 0.8 THz; Filter orders ( $N$ ) = 3, 4, and 5.
- Geometry Constraints: Photolithography limits required dimensions $>2\mu m$ . The center-to-center spacing ( $S+W$ ) was kept constant at $115\mu m$ .
- Impedance Selection:
  - Feedline ( $Z_0$ ): $234\Omega$ ( $S=70\mu m, W=45\mu m$ ).
  - High Impedance ( $Z_H$ ): $362\Omega$ ( $S=105\mu m, W=10\mu m$ ).
  - Low Impedance ( $Z_L$ ): $131\Omega$ ( $S=15\mu m, W=100\mu m$ ).
- Synthesis Calculation: Section lengths ( $l_n$ ) were calculated using standard synthesis equations relating element values ( $g_n$ ) from Bessel prototypes to the physical lengths of the stepped-impedance sections.
Experimental Setup:
- A modified THz-Time Domain Spectroscopy (THz-TDS) system was used.
- Transmitter: A low-temperature grown Gallium Arsenide (LT-GaAs) photoconductive switch bonded to the CPS TL, biased at 24V.
- Excitation: 80-femtosecond optical pulses from a mode-locked fiber laser.
- Receiver: Another LT-GaAs PCS with a mechanical delay line to reconstruct the temporal signal via lock-in detection.
- Range: The system characterizes devices from DC to beyond 3 THz.

3. Key Contributions

First Synthesized Filter in THz: This work represents the first demonstration of an integrated, planar, guided-wave synthesized filter (specifically a Bessel LPF) operating within the THz gap.
Validation of Microwave Synthesis: It proves that classical microwave filter synthesis methods (calculating geometry based on desired transfer functions) are directly applicable and valid at THz frequencies.
TSoC Platform Demonstration: It successfully utilizes a planar CPS TL platform with photoconductive switches to overcome the limitations of free-space optics and bulky electronic extenders, enabling broadband characterization up to 3 THz.
Design-Analysis Alignment: The paper bridges the gap between theoretical design and experimental reality in the THz regime, moving beyond simple periodic structure analysis to targeted filter synthesis.

4. Results

Temporal Response: The experimental time-domain data showed minimal pulse distortion for all filter orders ( $N=3, 4, 5$ ). This confirms the linear phase response inherent to Bessel filters, which is critical for transient systems to avoid ringing and preserve spectral resolution.
Spectral Response:
- The frequency domain response (obtained via Discrete Fourier Transform) showed a distinct change in slope (roll-off) near the designed cut-off frequency of 0.8 THz.
- The roll-off rate increased with the filter order ( $N$ ), matching theoretical expectations.
- The experimental data aligned closely with both theoretical predictions and full-wave simulations (ANSYS HFSS).
Performance: The filters successfully demonstrated the transition from pass-band to stop-band behavior consistent with the synthesized Bessel prototype.

5. Significance

Paradigm Shift: This work shifts the THz community from "design and analyze" (trial-and-error with metastructures) to "specify and synthesize" (targeted engineering of filter responses).
Enabling Technology: By validating synthesis methods, this paves the way for complex THz signal processing systems, including advanced multiplexing, noise reduction, and high-speed communication links.
Scalability: The methodology is not limited to CPS; the authors note that the synthesis approach is equally viable for other planar TL configurations (CPW, slotline), suggesting a broad applicability for future THz integrated circuits.
Cost-Effective Characterization: The use of the TSoC platform with photoconductive switches offers a relatively inexpensive and highly compatible method for characterizing THz devices compared to expensive electronic extension modules or complex free-space optical setups.