# Chip-Scale Terahertz Imaging for Security and Medicine

An episode of [The Hardware Podcast with Fexingo: Chips, Devices, and Electronics Engineering Conversations](https://invisiblenetwork.ai/podcast/1896799982), 11 Aug 2026, 9:48.

Audio: https://audio.fexingo.com/business/the-hardware-podcast/episode-0152.mp3

In this episode of The Hardware Podcast, Lucas and Luna explore how terahertz imaging is finally leaving the lab and shrinking onto a chip. They break down the physics of the 'terahertz gap,' the progress in silicon-based sources and detectors, and the two big use cases that are driving development: non-invasive security scanning and medical diagnostics. With specific examples like the Passport scanner and a 2025 Nature paper from MIT, they explain why this technology has been so hard to commercialize and what recent breakthroughs mean for the next few years. A practical, numbers-driven conversation about a sensor technology that could replace X-rays in some roles and make full-body scanners at airports faster and safer. #TerahertzImaging #ChipScaleSensors #SiliconPhotonics #SecurityScanning #MedicalImaging #MITResearch #NaturePaper #PassportScanner #THzTechnology #HardwareEngineering #Semiconductors #Sensors #NonInvasive #ImagingTechnology #TechPodcast #HardwarePodcast #FexingoBusiness #BusinessPodcast Keep every episode free: buymeacoffee.com/fexingo

## Transcript

Lucas: So the terahertz gap has always been this frustrating stretch of the electromagnetic spectrum — between microwaves and infrared — where we've known for decades there's huge potential, but actually generating and detecting that radiation reliably has been a slog.
Luna: Right, it's like the 'almost usable' band. You can see through packaging, but not with the resolution you'd want, and the equipment used to be the size of a lab bench.
Lucas: Exactly. But that's starting to change. Over the last few years, researchers have been finding ways to put terahertz sources and detectors onto silicon chips, and the implications are pretty big — especially for security scanning and medical diagnostics.
Luna: And that's what we're digging into today. How exactly do you shrink a terahertz system down to a chip, and what can it actually do that current tech can't?
Lucas: Let's start with the basics of the gap itself. Terahertz radiation sits roughly between three hundred gigahertz and three terahertz. It's non-ionizing, so it doesn't have the health risks of X-rays, but it can penetrate a lot of materials — plastics, fabrics, cardboard — so it's perfect for non-invasive inspection.
Luna: And the challenge has always been that it's too high-frequency for conventional electronics and too low-frequency for optics, so you're stuck in this no-man's-land.
Lucas: Right. Traditional terahertz sources were either giant free-electron lasers or bulky photoconductive antennas driven by femtosecond lasers. A whole setup could fill a room and cost millions. That's why it stayed in academic labs for so long.
Luna: So what changed? How are we getting it onto a chip?
Lucas: The key breakthrough has been in silicon-based technologies. There are a couple of different approaches, but the one that's getting a lot of attention is using cmos compatible processes to build terahertz sources and detectors — essentially the same manufacturing line you'd use for a smartphone processor.
Luna: So we're talking about leveraging the entire semiconductor industry's economies of scale to make terahertz devices mass-producible.
Lucas: Exactly. One of the most cited examples is from a few years back at MIT, where they built a chip that can generate terahertz waves using a conventional CMOS process. That chip, which they demonstrated in a paper, was small enough to fit on a fingertip and could produce signals in the terahertz range.
Luna: And the applications they were talking about were things like security scanning and even medical imaging.
Lucas: Right. And since then, we've seen more groups pushing in that direction. There's also a Finnish company, VTT, that has been working on terahertz sensors for industrial quality control. But the big, high-profile use case is security scanning — like what you see at airports.
Luna: Because terahertz can see through clothing without being ionizing, so you get a scan that's less invasive in terms of privacy and health.
Lucas: And actually, there's a company called Passport that makes a scanner based on terahertz technology — it uses those non-ionizing waves to detect concealed objects, and it's already used in some security settings. But those systems are still fairly large and expensive. The promise of chip-scale terahertz is that you could shrink that whole scanner down to something portable, maybe even handheld.
Luna: That would be a game-changer for security checkpoints, but also for things like finding hidden objects in walls or detecting explosives from a safe distance.
Lucas: And then there's the medical side. Terahertz radiation is very sensitive to water content, so it can distinguish between healthy tissue and cancerous tissue, which has different water concentrations. Skin cancer detection is a big one — a chip-scale terahertz sensor could let dermatologists scan suspicious moles in real time without needing a biopsy.
Luna: So it's like a non-invasive, real-time pathology tool. That's pretty compelling.
Lucas: Now, before we get deeper into the technical hurdles, we should mention something. This show, like many things, runs on listener support. A couple of dollars a month is genuinely what keeps these episodes coming out ad-free. If you've gotten something out of our hardware conversations, you can help at buy me a coffee dot com slash fexingo. It really does make a difference.
Luna: Yeah, and there's no paywall or premium tier — it's just people who appreciate the work keeping it alive.
Lucas: Anyway, back to the chips. The main technical problem is that silicon isn't naturally good at emitting or detecting terahertz waves. Silicon is transparent to terahertz in some ways, but generating it efficiently has been tough. The chip from MIT used a technique where you generate terahertz by mixing two lasers on the chip, but that's not trivial to do in a compact way.
Luna: And the detectors are also tricky. You need something that can sense very weak signals, and at terahertz frequencies, thermal noise is a real issue.
Lucas: Right. So there are trade-offs. One approach is to use what's called a Schottky diode, which can rectify terahertz signals into a DC voltage. Those have been around for a while, but integrating them on a chip with the antenna is the trick.
Luna: And you also have to worry about the power output. Terahertz sources on chip tend to be low-power, which limits how far the signal can travel and how deep it can penetrate.
Lucas: Yeah, but there's been progress. In 2025, a team at MIT published a paper in Nature demonstrating a chip-scale terahertz source that could produce over a milliwatt of power — that's a big deal because it's enough for practical sensing applications.
Luna: A milliwatt might not sound like much, but in the terahertz world, that's a leap.
Lucas: Exactly. And it's using a standard CMOS process, so it could be produced at scale. The paper also showed that the chip could be used for high-resolution imaging, with a resolution of about a millimeter, which means it could see small objects clearly.
Luna: So we're at the point where the pieces are coming together. But what's the bottleneck for commercialization now?
Lucas: I'd say it's a combination of integration and cost. You need the source, the detector, the antenna, and the readout electronics all on a single chip, and that's not easy to get right. Plus, there's the question of how you package it and whether it can survive real-world conditions.
Luna: And there's also the regulatory side, especially for medical devices. You can't just slap a terahertz sensor on a dermatologist's desk without going through clinical trials.
Lucas: Right. But the security side might move faster. There's less regulation for non-medical devices, and the need is clear — think about screening at stadiums or public events without causing lines.
Luna: I can see a future where a security guard waves a handheld terahertz scanner over a crowd and picks up concealed weapons from a distance.
Lucas: That's the vision, and it's closer than you might think. A few startups are already working on portable terahertz scanners, though they're not yet at the chip-scale level. But given the pace of research, I'd say we're five to ten years away from seeing a truly chip-based terahertz imager in commercial products.
Luna: That's a realistic timeline. And with the potential to replace X-rays in some applications, it's definitely a space to watch.
Lucas: Yeah, and the interesting thing is that this is happening across multiple fronts — not just imaging but also spectroscopy. Terahertz waves interact with molecular vibrations, so you can identify chemicals by their terahertz signature. That opens up the door for detecting drugs, explosives, or even counterfeit pharmaceuticals.
Luna: So it's not just about seeing shapes; it's about identifying substances.
Lucas: Exactly. And that's why chip-scale terahertz is so exciting — it could bring that capability out of the lab and into the field. A customs officer could scan a package and know what's inside without opening it.
Luna: That would be a huge win for security and public safety.
Lucas: Of course, there are still challenges. The power levels are still relatively low, so the range is limited to a few meters, but that's fine for close-range inspection. And the resolution, while improving, isn't yet at the level of X-rays for some applications.
Luna: But for many use cases, you don't need that high resolution. You just need to know if there's something hidden or if a sample is cancerous.
Lucas: Right. And the fact that it's non-ionizing is a big plus. We're moving toward a world where we don't have to irradiate people to screen them, and that's a meaningful change.
Luna: So, what's the takeaway? Are we on the cusp of a terahertz revolution?
Lucas: I think we're on the cusp, but the revolution will be quiet. It won't be a single breakthrough; it'll be a series of incremental improvements that eventually make chip-scale terahertz viable. And when that happens, it'll be everywhere — in security, medicine, industrial quality control, even in your phone for gesture recognition.
Luna: Gesture recognition? That's a wild thought.
Lucas: It's been demonstrated in the lab, but it's not ready for prime time. But the point is, the potential is enormous, and the semiconductor industry is finally paying attention. I'd keep an eye on the next few years — this is one of those technologies that could suddenly become ubiquitous.
Luna: I'm looking forward to that. Thanks for breaking it down, Lucas.
Lucas: Anytime. And if you're curious, the MIT paper is a good place to start — it's a great example of what's possible.

Shows are listed from Apple Podcasts' charts and directory, and Podcast Index where it is used. Every show and episode belongs to its publisher and plays from the publisher's own feed; this site hosts none of them.