Terahertz waves are coming to the real world

Terahertz (THz) waves, commonly understood to correspond to frequencies from approximately 0.1 to 10THz, interact with the vibrational resonances of many molecules. This results in absorption or radiation at specific frequencies. Therefore, THz waves have long been investigated in astronomy and spectroscopy for identification of molecules in space and to characterize the composition of matter, respectively. Several technical breakthroughs made over the last few decades now allow us to generate and detect THz waves more easily. This has triggered a search for new uses of the technology in many fields, including bioscience, security, and information and communications technology.¹ Recent advances include reliable time-domain spectroscopy, achieved by combining a femtosecond laser and photoconductive switches, and medical imaging and nondestructive testing systems whose THz-wave signal sources are based on nonlinear optical effects or use advanced semiconductor quantum devices. However, the complex technologies and bulky equipment are not suitable for practical use, especially in outdoor sensing and wireless-communications applications.

To develop compact and reliable THz-wave applications, we are using photonic technologies, which were originally developed for fiber-optic communications. One of the key photodetectors resulting from our work is a uni-traveling-carrier photodiode (UTC-PD: see Figure 1). Its photocurrent is dominated by a uni-carrier (specifically electrons, which are much faster than holes), resulting in very fast photo response.^2,3 This is the main difference between our new device and conventional p-i-n photodetectors, where both electrons and holes contribute equally. Since their invention in 1997, UTC-PDs have been optimized to achieve higher output power at THz-wave frequencies. Recent, advanced devices exhibit promising performance for THz-wave applications. The output power has reached almost 0.5mW at 350GHz, which is sufficient for our short-distance (up to a few tens of meters) applications, such as remote sensing and ultrahigh-capacity wireless communications at frequency ranges from 200 to 500GHz, where atmospheric attenuation is relatively low.

Figure 1. (a) J-band (10–20GHz) module with horn antenna. (b) Uni-traveling-carrier photodiode (UTC-PD) with log-periodic antenna. (c) UTC-PD with dipole antenna. RF: Radio frequency. PAD: Bias pad.

Using UTC-PDs, we first developed a THz-wave signal generator for remote sensing. To maintain the highest possible signal-to-noise ratio, we had to generate a monochromatic THz-wave signal with a very narrow linewidth to increase the spectral density for a given power. Our signal generator consists of a UTC-PD, arrayed-waveguide gratings, optical switches, and an optical comb-signal generator equipped with nonlinear fibers.⁴ The latter exhibits excellent coherency between modes, enabling a very narrow linewidth of the output signal of a few hertz at 300GHz. Its phase noise is as low as that of instrumental-grade microwave-signal sources. Another key feature is our system's wide frequency tunability in the range from 100GHz to 1THz, which can be tuned continuously or even randomly with kHz-order resolution, while the output frequency can be locked to other frequency references. This means that phase information can be extracted using both homo- and heterodyne detection. Using the THz-wave signal generator, we have demonstrated a simple spectroscopy system for gas identification with a Schottky-barrier-diode detector. Although the test's frequency span was not very wide because of the waveguide packaging of the emitter and detector, the nitrogen dioxide's absorption characteristics were clearly obtained within a scanning time of 1min (see Figure 2).

Figure 2. Measured (scatter) and calculated (solid line) transmittance (in arbitrary units, a.u.) of nitrogen dioxide gas.

We also investigated THz waves for wireless communications. Because they offer extremely wide bandwidth (more than 100 times wider than that of conventional cellular systems), data capacities of up to 100GB/s are expected.⁵ The frequency band in excess of 275GHz, which has not yet been allocated for specific use, is especially attractive for this purpose. When 100Gbps wireless links are finally realized, one will be able to download a Blu-ray^® movie (approximately 25GB) to a memory card embedded in a smart phone in just a second without having to take the phone out of one's bag or pocket. For this communications application, photonic technologies are advantageous compared to electronic approaches because of their inherent broadband nature. Photonic technologies can generate a high-frequency carrier signal and are also able to handle extremely broadband data signals. We recently performed a preliminary data-transmission experiment,⁶ in which 12.5GbB/s data was carried on THz waves at 300GHz and transmitted over a 50cm-long distance without errors. Figure 3 shows measured bit-error rates at several data rates and an eye diagram for a rate of 12.5GB/s. We also used the UTC-PD as a transmitter and performed amplitude-shift-keying data modulation with a commercial optical-intensity modulator. Taking the performance margins of the transmitter and receiver (such as signal power and receiver bandwidth) into consideration, we believe that data can be transmitted at up to 20GB/s.

Figure 3. Measured bit-error rate of the terahertz-wave wireless links at several data rates and eye diagram (power as a function of time) for a rate of 12.5GB/s (Gbps).

Photonic technologies developed for telecommunications systems can now play an important role in practical THz-wave applications, such as remote sensing and wireless communications. UTC-PDs can produce a wave at up to 1THz, and many other optical components for 1.55μm telecommunications systems enable easy implementation of a variety of functions at low cost. Once we improve the output power of our UTC-PDs, sensing over longer distances or transmission of much larger data volumes will become possible.