Making waves

Introducing the terahertz range

The generally accepted definition of the terahertz (THz) range is, in frequency units, 300 GHz, ie 0.3THz, up to 10THz. In wavelength terms this corresponds to 1mm to 30μm.

THz radiation has many useful properties. It passes easily through materials, such as paper, packaging, wood, ceramics, certain building materials, and plastics. It will also travel a few millimetres in human tissue and up to 500m in the atmosphere. With the right type of equipment, a THz map or image of a specimen can be built up.

Terahertz waves are, however, notoriously difficult to produce and detect. ‘THz is the most challenging radiation in the whole electromagnetic spectrum to generate,’ says Daniel Mittleman of Rice University, Texas, US. ‘Taking low-frequency generation and scaling it up gets increasingly less efficient. But taking high-frequency generation and scaling it down, also gets increasingly less efficient. There is a ‘gap’ in the middle of the range at around 1 THz.’

Black body sources, such as filament lamps, can be heated to generate infrared radiation. ‘But emissions roll off exponentially as wavelength increases and, at the longer wavelengths, there is not much radiation at all,’ explains Mittleman. From the other end, microwave technology can be scaled up in frequency. ‘But to produce THz radiation, oscillations have to be so fast that the radiation efficiency of the antenna diminishes’, he says.

Over the years, there have been many attempts and different techniques applied to the ‘THz gap’. ‘The problem is how to conjure a way to generate THz and get it the hell out of the device without being absorbed,’ says Eric Mueller of Coherent, a laser manufacturer in the US. ‘The vast majority of systems work, but they are weak, producing nano- to micro-watts of power on average.’

Generating THz radiation

One way of generating THz radiation uses optically pumped lasers. Light is used to raise or ‘pump’ electrons from a lower energy level in an atom or molecule, to a higher one. In an optically pumped terahertz laser (OPTL), the laser consists of a vacuum cell containing a molecular gas – usually carbon dioxide – at low pressure; some source of optical feedback (end mirrors); and a method of shinning infrared radiation and emitting THz radiation. 

‘It’s not as easy as it sounds,’ says Mueller. ‘Slight changes in the pumping wavelength of the CO₂ laser or changes in the length of the cell can inhibit production of THz radiation, and a feedback interaction between the pump laser and the terahertz laser can affect stability of the system.’

In the past, research groups often built their own OPTLs, which were typically large and extremely difficult to use and maintain, he says. Today, with improvements in laser technology and sealed gas cells, OPTL laser systems are smaller and more reliable. OPTLs can operate at many discrete frequencies, ranging from less than 300 GHz (1000 µm) to more than 10THz (30µm), depending on the gas used. Each gas has its own spectrum of available lines that can be used to produce THz radiation at specific frequencies.

OPTL lasers are in use around the world, primarily for astronomy, environmental monitoring, and plasma diagnostics. One 2.5THz laser, made by Coherent, is currently onboard a NASA satellite, measuring levels of hydroxyl radical in the stratosphere. However, as Mueller says, they are expensive. ‘The technology is reliable and proven, and power output is good, between 10-100microwatts. But there is no compelling application to drive further development.’

Another area of interest is time domain systems. Here, an optical laser pulse (100 femtoseconds or shorter) creates carriers (electron-hole pairs) in a semiconductor material. The semiconductor changes abruptly from an insulator into a conductor, which results in a sudden electrical current across an antenna, patterned on the semiconductor. The changing current emits THz radiation, similar to what happens in the antenna of a radio transmitter.

The short duration of THz pulses generated (typically less than 1 picosecond) is primarily a result of the rapid rise of the photo-induced current in the semiconductor and the short carrier life-time of semiconductor materials. This current may persist for only a few hundred femtoseconds, up to several nanoseconds, depending on the material.

According to Mittleman, this technology is growing rapidly. ‘There is a big push here. These systems operate over very broad bandwidth and cannot achieve a high resolution, which may be necessary for some applications. And, technologically, they require a femtosecond laser, which is a very expensive piece of equipment. They need to be smaller and cheaper if they are to gain widespread use in the world.’ Teraview, a company based in Cambridge, UK, is one of several developing this technology.

A new twist on time-domain systems has emerged from a joint institute of the US National Institute of Standards and Technology and the University of Colorado at Boulder. Lead researcher Steven Cundiff claims their system is unusually efficient.¹ Their innovative approach involves adding a layer of silicon oxide insulation between the semiconductor and gold electrodes, used to apply a field across the semiconductor, which prevents sample damage.

‘Users can boost THz power by raising the optical power without damaging the semiconductor,’ says Cundiff. ‘Among other advantages, the new technique does not require a microscopically patterned sample or high-voltage electronics. The system produces a peak terahertz field comparable to that of other methods.’

Another area of effort is direct multiplied (DM) sources. The idea is to take millimetre-wave sources and directly multiply their output up to THz frequencies. Mueller explains, ‘DM sources with frequencies up to a little more than 1THz and approximately 1µW of output have been used in select applications, most of which are in radio astronomy. However, they can produce substantially more power at lower frequencies, and they are often well suited to applications requiring frequencies of less than 500GHz.’

Meanwhile, other researchers have been investigating quantum-cascade semiconductor lasers, which operate at wavelengths in the 1.3-4.4THz region. These lasers are made from hundreds of alternating layers of semiconductors, such as gallium arsenide and aluminium gallium arsenide. Applying a potential across the device causes electrons to cascade through each level, emitting photons along the way. The photon wavelength is determined by the thickness of the stages. ‘You can tune the radiation [to THz] by changing the thickness of the various layers,’ says Mittleman.

However, DM systems only operate well at cryogenic temperatures. This makes them inconvenient, expensive, limits market use and presents safety issues. But as Mittleman points out, ‘operating temperatures are creeping up all the time’, and ultimately they do offer the potential to be small, compact and cheap because they are based on semiconductor lasers, which can be mass-produced.

Mueller confirms the main challenge is that while they can produce substantial output power at liquid helium temperatures, they barely work at liquid nitrogen temperatures. It is not clear whether this technology will ever support room-temperature operation, he says, but, as its operating point is being steadily increased by researchers, its potential usefulness for many applications will inevitably increase, he adds.

There are many different technologies that can generate THz. Choosing which one will depends on its application. Mittleman explains, ‘For example, if you want to detect how much hydrogen sulfide is coming out of a smokestack, then a narrowband system tuned to H2S frequency would be fine, such as a quantum-cascade laser or a direct multiplied source, depending on the frequency you wanted to hit. But if you wanted to look under a bandage to see if a wound is healing, then you would be better off with a broadband source such as a time domain source, which lends itself to imaging configurations with high depth resolution.’

Challenges in detection

The same reasons that make THz generation difficult, also make them difficult to detect. Indeed, detection can be more challenging, says Xi-Cheng Zhang of the Rensselaer Polytechnic Institute (RPI) in New York, US.

Naturally occurring moisture in the air absorbs broadband THz waves, explains Zhang. This means that detection must be done under dry nitrogen-purged environments or at short distances to avoid loss of signal strength (several centimetres). He continues: ‘In addition, since the electromagnetic transients [radiation] last for only picoseconds in time (10–12s), an optical pulse must be used to sample the field. Aligning the THz radiation and its sampling optical pulse is critical. What’s more, many of the materials used for detection are limited in bandwidth due to their crystal structure.’

Ben Clough of the RPI, working with Zhang, has come up with a patent-pending solution to extend the effective distances over which THz could be detected². His method uses sound waves to ‘listen’ to THz signals from a distance.

Focusing two laser beams into air creates small bursts of plasma, which in turn create THz pulses. Another pair of lasers is aimed near the target to create a second plasma for detecting the THz pulses after they have interacted with the material. This detection plasma produces acoustic waves as it ionises the air.

Clough discovered that by using a sensitive microphone to ‘listen’ to the plasma, he could detect THz wave information embedded in these sound waves. This audio information can then be converted into digital data and instantly checked against a library of known THz fingerprints, to determine the chemical composition of a material.

The RPI team says it has demonstrated remote THz detection at up to 10m and remote generation at up to 30m separately in two individual experiments. They are working towards remote spectroscopy by combining these techniques.

The approach is novel because of the use of acoustic waves to sense radiation lasting only picoseconds, says Zhang. ‘It also allows detection of THz radiation in any direction, unlike other methods that require collection of the signal in the forward propagating direction. It is a relatively cheap detection method, in comparison with optical methods, as it does not require expensive optics and critical alignment. It also does not require direct line of sight to the plasma sensor like all other optical techniques, making transmission of the information through optically opaque barriers feasible.’

The main challenges remaining in developing any commercial detection system are sensitivity, selectivity, simplicity, scalability and stability, according to Zhang.

Applications

Imaging specialist Teraview has been developing THz systems to detect skin, mouth and throat cancers in the field of medical imaging. Such systems could also be used to detect tooth decay at an early stage and monitor early erosion.

The problem with medical applications, says Mueller, is that regulatory authorities require a tremendous level of certainty. ‘Nobody has shown the technology works with a large enough number of patients to generate acceptable accuracy rates’. Also THz radiation can’t penetrate too far into the body, so applications are limited to detecting surface cancers or as probes during surgery.

Meanwhile, security and defence-related applications remain in the R&D stage, Zhang says. Even so, Mueller is unconvinced whether this area will ever prove viable. ‘To have a THz system that truly works in the real world, that is able to interact with different materials and scenarios, is very challenging.’

However, Mueller believes that there are potential applications yet to be developed. ‘Getting the first volume application will change the environment significantly. Most likely, this breakthrough will be in non-destructive testing, where imaging techniques can look at manufacturing problems in products before they are shipped,’ he predicts. THz can be sent through materials in pulses, making it possible to detect foreign bodies or flaws.

Zhang also agrees that, in the short-term, the most commercially successful application will be industrial non-destructive testing (NDT). ‘[We have had] more than 80% requests from industrial companies interested in NDT applications, including defects detection, and also for pharmaceutical purposes, such as monitoring drug concentrations.’