Every generation of mobile technology has been a story about climbing higher up the radio spectrum. As each band fills with traffic, engineers reach for the next, less crowded stretch of higher frequencies, trading easy range for greater capacity. The coming sixth generation is poised to make the boldest leap yet, into a region of the spectrum so difficult to use that physicists long called it a gap. This is the terahertz band, and understanding both why it is so attractive and why it is so stubborn is the key to understanding what 6G can and cannot become.
The last unused stretch of spectrum
The terahertz band sits between microwaves and infrared light, conventionally spanning from about 0.1 to 10 terahertz. For decades it was the least exploited part of the electromagnetic spectrum, too high for traditional electronics and too low for optical devices, which is exactly why it earned the nickname the terahertz gap. Generating and detecting these waves efficiently was simply beyond the reach of available hardware.
That neglect is now its greatest appeal. While the millimeter-wave bands used in advanced 5G can rarely offer system bandwidths above a single gigahertz, the terahertz region opens up vast, untouched swaths of spectrum, with contiguous bandwidths potentially exceeding fifty gigahertz. In wireless communication, bandwidth translates almost directly into speed, and this is why the band has become central to the 6G vision.
Why 6G needs it
The headline ambition for 6G is a peak data rate of one terabit per second, roughly a hundredfold jump over the targets of the previous generation. Demand for data keeps climbing, and the existing microwave and millimeter-wave channels cannot stretch to meet it. Only the enormous bandwidths of the terahertz band make terabit links physically conceivable.
Speed is not the whole story. The same physical properties that make terahertz waves fast also make them exquisitely sensitive instruments. Their short wavelengths give fine spatial resolution, and their extremely short pulses, measured in picoseconds, allow precise measurement in time. This enables a feature the industry calls integrated sensing and communication, in which a single system both carries data and senses its surroundings, mapping a room or tracking movement while it transmits. A terahertz network could double as a high-resolution radar, supporting applications from holographic telepresence to centimeter-accurate positioning. It is this fusion of communication and sensing, as much as raw speed, that makes the band so compelling.
The tyranny of attenuation
If terahertz were merely a matter of faster speeds, it would already be everywhere. The reason it is not comes down to a brutal physical reality: these waves do not travel well. The problem is often described as a dual attenuation dilemma, and both halves of it are severe.
The first is free-space path loss, the natural weakening of any signal as it spreads outward, which grows dramatically worse at higher frequencies. At three hundred gigahertz, a signal can lose around a hundred decibels over just ten meters, and across a kilometer the loss exceeds a hundred and twenty decibels, far beyond anything seen in millimeter-wave systems. The second is atmospheric absorption. Water vapor and oxygen molecules in the air resonate at terahertz frequencies and greedily absorb the energy passing through them. Push above one terahertz and the air itself becomes nearly opaque, capable of weakening a signal tenfold within a single meter. Rain and fog make it dramatically worse.
The practical consequence is that terahertz communication is inherently short-range. These are not waves that will blanket a city from a distant tower. They are waves for closing small distances at colossal speed.
Finding the windows
The situation is not hopeless, because the absorption is not uniform across the band. Between the peaks where molecules resonate most fiercely lie relative calm zones known as atmospheric windows, at frequencies such as one hundred forty, two hundred twenty, and three hundred forty gigahertz. Within these windows the molecular absorption drops to only a couple of decibels per kilometer, almost negligible next to the free-space loss. Much of the early engineering work, including outdoor measurement campaigns in dense urban settings, has focused on the sub-terahertz region around one hundred forty gigahertz precisely because it offers a usable window for realistic distances.
Even inside these windows, the path loss demands clever compensation. Systems must concentrate their energy into tightly focused, highly directional beams using high-gain antennas and beamforming, rather than broadcasting in all directions. A further tool is the intelligent reflecting surface, a programmable panel that can steer a beam around obstacles and toward a specific user, helping these fragile signals navigate a cluttered real world. Other stubborn problems remain, including dispersion effects that smear signals in time and the fact that existing modulation and coding schemes are not well suited to such enormous bandwidths.
A technology still being born
It is worth being honest about how early this all is. The hardware needed to generate and detect terahertz waves cheaply and at scale is still immature compared with the refined electronics of lower bands. Realizing the vision will require not just better components but a dense deployment of base stations to compensate for the limited range, a significant and costly undertaking.
The standardization effort reflects this measured pace. International bodies have agreed on a framework and timeline for 6G, with the years immediately ahead devoted to defining requirements and evaluation criteria, and formal technical standardization of terahertz systems expected to begin only afterward. Coordinating global spectrum allocation and ensuring devices work across regions is a parallel challenge that governments, industry, and academia must solve together.
A realistic horizon
Terahertz radiophysics offers a genuine path past the looming wall of spectrum scarcity, promising data rates and sensing abilities that lower frequencies can never deliver. But physics imposes hard limits that no amount of engineering will fully erase. The most likely outcome is not a wholesale replacement of today’s networks but a layered one, where terahertz links handle short-range, ultra-high-capacity tasks, such as the wireless backbone between base stations, instant transfers in a single room, or dense indoor hotspots, while lower frequencies continue to provide the wide coverage they always have. The terahertz gap is finally being bridged, but it will be crossed deliberately, one atmospheric window at a time, by those who respect both its extraordinary promise and its unforgiving constraints.