Inexpensive detector for terahertz imaging A. Abramovich,1,* N. S. Kopeika,2 D. Rozban,2 and E. Farber2 1
Department of Electrical and Electronic Engineering, Ariel University Center of Samaria, Ariel, Israel Department of Electro-Optical Engineering, Ben Gurion University of the Negev, Beer-Sheva, Israel
2
*Corresponding author:
[email protected] Received 2 May 2007; revised 30 July 2007; accepted 30 July 2007; posted 8 August 2007 (Doc. ID 82639); published 5 October 2007
Glow discharge plasma, derived from direct-current gas breakdown, is investigated in order to realize an inexpensive terahertz (THz) room-temperature detector. Preliminary results for THz radiation show that glow discharge indicator lamps as room-temperature detectors yield good responsivity and noiseequivalent power. Development of a focal plane array (FPA) using such devices as detectors is advantageous since the cost of a glow discharge detector is approximately $0.2–$0.5 per lamp, and the FPA images will be diffraction limited. The detection mechanism of the glow discharge detector is found to be the enhanced diffusion current, which causes the glow discharge detector bias current to decrease when exposed to THz radiation. © 2007 Optical Society of America OCIS codes: 110.6795, 120.0280, 120.1880, 040.2235, 040.1880, 020.2070.
1. Introduction
The electromagnetic spectrum between 3 mm and 30 m has become attractive for applications in medicine, communications, homeland security, and space technology. This is because there is no known ionization hazard for biological tissue, and Rayleigh scattering of terahertz (THz) radiation is low compared with infrared and optical rays. The lack of inexpensive room-temperature detectors in this spectral region makes it difficult to develop detection and imaging systems. The interaction between microwave and glow discharge tubes has been investigated theoretically and experimentally [1–3]. The use of glow discharges as electromagnetic radiation detectors is advantageous due to their low cost, wide dynamic range, electronic ruggedness, broad spectral range, room-temperature operation, and simplicity of use [3]. Glow discharge detectors (GDD) based on plasma discharge have served as video detectors [4 – 6] and synchronous detectors [7]. The change in drift current Id of the GDD in the case of a large dc field and a small ac electric field is [6]
0003-6935/07/297207-05$15.00/0 © 2007 Optical Society of America
⌬Id ⫽ 共⌬i兲n ⫹ i共⌬n兲 ⫺ 共⌬D兲ⵜ2n ⫺ Dⵜ2共⌬n兲, (1) where i is the ionization collision frequency, n is the charge-carrier volume density, and D is the diffusion coefficient. The influence of the THz field can be expressed in terms of an increase in ionization current rate or diffusion current. The change in n is a function of position x between the GDD electrodes. The total change is the integration of all regions between and around the electrodes. The use of GDDs was limited since they have internal noise, which is a characteristic of plasma devices [2,4]. Nevertheless, reduction of the noise levels of GDDs [5,8] using sophisticated methods can allow them to be used advantageously in more widespread applications. In this paper we demonstrate the use of a GDD in the millimeter-wave and terahertz (THz) frequency radiation region. We present experimental results in which we were able to detect 100 and 250 GHz radiation using commercial indicator lamp GDDs. The THz spectral region is one with many possible applications, presently limited to a large extent by a relative lack of inexpensive but sensitive detectors. We hope that GDD devices can help alleviate this problem. In this work, we found that the dominant GDD detection mechanism in this region is the enhanced diffusion current, which causes the 10 October 2007 兾 Vol. 46, No. 29 兾 APPLIED OPTICS
7207
GDD current to decrease, in contrast to the detection mechanism in microwave radiation detection experiments [6], which was attributed to rf enhanced ionization [3], which increased the bias current. 2. Experimental Setup and Results
There are two THz sources that we used to investigate the GDD. The first is for 100 GHz and the second is for 250 GHz. The 100 GHz experimental setup is based on GaAs multipliers from Virginia Diodes, Inc. (Charlottesville, Virginia, USA) [9] that multiply a low-frequency source to 100 GHz. In our case we use an ordinary 0–18 GHz synthesizer as a lowfrequency source and multiply it by 8 to obtain 92–102 GHz. The power levels of this radiation are 150–200 mW, depending on the required frequency. The second THz source of 250 GHz is based on a backward wave oscillator (BWO), which is a part of a THz quasi-optic system [10]. In this experimental setup we used a BWO (catalog no. OV 24N123) operating at 173–260 GHz, which provides 20–30 mW, depending on the frequency. Both sources radiate to free space, and via a polyethylene (PE) lens the THz radiation was focused on
the GDD cross section between the electrodes. The basic experimental setup is given in Fig. 1. The 12.5 GHz signal of the rf synthesizer was modulated with a 1 kHz square wave. This modulated signal was used to derive the ⫻8 GaAs multiplier. The 100 GHz signal of the multiplier was coupled to free space by a rectangular horn antenna and produces an approximately fundamental mode Gaussian beam [11,12]. The THz Gaussian beam was focused by a PE lens on the GDD. The GDD was operated in the abnormal glow region of the current voltage characteristics [3], and it was connected to an amplifier and to a scope. Since the modulation of the THz beam was 1 kHz we tuned the bandwidth of the amplifier to be between 100 and 10,000 Hz in order to reduce the noise of the detection system. The output signal was obtained on the scope and was recorded by a special computer code. A commercial green neon indicator lamp, N523 shown in Fig. 1(a), from Internal Light Technologies (Peabody, Massachusetts, USA), was used in these experiments as a GDD. The N523 showed the best performance as a detector among neon indicator lamps from several companies that were examined using the basic setup of Fig. 1(b).
Fig. 1. (a) Description of the GDD N523 (measurements are in mm). (b) Basic experimental setup for evaluating the GDD performance as a THz detector. 7208
APPLIED OPTICS 兾 Vol. 46, No. 29 兾 10 October 2007
Fig. 2. GDD signal as obtained on the scope in arbitrary units as a function of time 共500 s兾div兲. The THz radiation was modulated with a 1 kHz square wave.
Fig. 4. Response of the GDD as a function of the GDD dc current. The output signal is in arbitrary units (the THz frequency is 250 GHz, and the optical chopper modulation is 700 Hz).
The detected signal of the GDD N523 is given in Fig. 2. This signal was obtained under the following conditions: amplification of 10,000, amplifier bandwidth of 100–10,000 Hz, GDD dc current of 3 mA, square-wave modulation at 1000 Hz, and THz frequency of 100 GHz. Moreover, it was found that the responsivity of the GDD varies with the dc bias current. Figure 3 shows that the higher the dc current of the GDD, the better the responsivity of the GDD. This is attributed to an increase of the plasma frequency with the bias current [3,5], thus increasing the THz radiation absorption. Control of the dc current in the GDD was performed by changing its voltage between 100 and 150 V. We repeated the responsivity experiment for a THz frequency of 250 GHz, but with about an order of magnitude less incident electromagnetic (EM) wave power. Figure 4 shows the GDD response at 250 GHz. In this experiment the modulation of the THz radiation was carried out with an optical chopper operating at approximately 700 Hz. An investigation of the GDD detection mechanism in the THz band was carried out. The detection mechanism at microwave frequencies several decades ago with older-model lamps was believed to be the enhanced cascade ionization that caused the GDD current to increase when exposed to rf radiation [3,6]. In
contrast to this assumption, the experimental results here with newer GDD devices show that the GDD current decreases when exposed to incident THz radiation. Figure 5 shows the square-wave modulation signal and the detected signal by the GDD on the same time axis. From Fig. 5 it is clear that when THz radiation is incident on the GDD the dc decreases. Thus, the detection mechanism here is not the enhanced cascade ionization. However, it was possible that the external amplifier might have induced 180° phase change or that the modulation signal might have 180° phase change caused by the source multiplier system. The former possibility was eliminated by increasing the received THz power so that the amplifier was not needed. Therefore, to eliminate possible 180° phase effects caused by the source multiplier system we considered whether the detected signal is manifested as a discharge current increase or decrease in a different type of experiment. Figure 6 describes this special setup. In this setup we used a HeNe laser beam to serve as our reference signal for the modulation. In other words, when the laser beam hits the laser detector located after the chopper, the THz radiation hits the GDD. When the laser beam is not blocked by the chopper, the laser detector signal increases. Figure 7 shows the HeNe laser signal and the GDD detected signal on the same time axis. This result completely agrees with the results of Fig. 5 and indi-
Fig. 3. Response of the GDD as function of GDD dc bias current. The signal is in arbitrary units (THz frequency is 100 GHz). We repeated the responsivity experiment for a THz frequency of 250 GHz.
Fig. 5. GDD detected signal and modulation signal of the THz radiation on the same time axis. 10 October 2007 兾 Vol. 46, No. 29 兾 APPLIED OPTICS
7209
Fig. 6. Experimental setup for investigating the GDD detection mechanism.
known to arise [6]. The negative space charge there and in other locations enhances the deflection in the electron trajectory to the anode, thus increasing the diffusion at the expense of the direct current to the anode. This reduces the bias current. Such a phenomenon was observed previously [6,13]. Furthermore, these lamps, unlike those used 35 years ago, have a phosphor layer on the inside surface of the glass. The THz enhanced diffusion current is likely to reach the glass surface, unlike a THz enhanced ionization current, which would go directly to the anode. Green photons released by the phosphor layer upon THz detection electron impact with it can then be absorbed by excited atoms, thus photionizing them or further photoexciting them for subsequent ionization collisions. Either way provides significant internal amplification of the THz detection signal. In other words, there are two detection mechanisms, but the phosphor layer on the glass strengthens the diffusion current mechanism to make it dominant over the ionization detection mechanism [see Eq. (1)]. The responsivity of the GDD was measured using a THz absolute power meter from Thomas Keating (Billingshurt, UK). A THz beam was focused on the GDD cross section. The size of the Gaussian beam was measured using a Spiricom (Logan, Utah, USA) THz camera. The responsivity curve of the GDD for a dc current of 9.5 mA is given in Fig. 8. The amplified 共30 dB兲 noise levels of the detection system were measured with a spectrum analyzer and found to be about 10⫺10 W at 1 kHz modulation over a 100–10,000 Hz electronic bandwidth (BW). Figure 9 shows the noise level of the detection system where the BW of the amplifier is 10 Hz to 1 MHz. A calculation of the GDD noise-equivalent power (NEP) was carried out using Eq. (2):
cates that the GDD detected signal is negative. The small mismatch in Fig. 7 is due to the difference in the HeNe beam waist and the THz beam waist on the chopper aperture plane. Experiments with these newer-model lamps at 10 GHz yield similar results. The detection mechanism here therefore appears to be the THz radiation enhanced diffusion current (negative current) instead of the enhanced ionization current (positive current). The former seems to occur in the Faraday dark area location of the GDD where the electrons have low energy, most of the negative space-charge gradient is concentrated, and a diffusion current is
where N is the noise of the GDD detection system, Vn is the voltage noise, and R is the responsivity of the GDD. The amplified responsivity according to Fig. 8 is approximately 20 V兾W, and the half-maximum electronic bandwidth in the 100–10,000 Hz filter
Fig. 7. GDD detected signal and modulation signal of the THz radiation on the same time axis for the chopper experimental setup.
Fig. 8. Responsivity curve of GDD N523 for a dc current of 9.5 mA. The THz frequency is 100 GHz modulated by 1 kHz, and the bandwidth of the amplifier is 100–10,000 Hz. The amplifier gain is 30 dB.
7210
APPLIED OPTICS 兾 Vol. 46, No. 29 兾 10 October 2007
N ⬀ Vn2 ⫽ 共R ⫻ NEP兲2 ⫻ BW,
(2)
Fig. 9. Noise spectrum of the GDD detection system as was measured by a spectrum analyzer where the BW of the amplifier is 10 Hz to 1 MHz. The amplifier gain is 30 dB.
band was measured to be only 1080 Hz. Thus, the NEP of the system was calculated to be 10⫺8 W兾冑Hz. If a higher modulation rate, such as 100 kHz, will be used, the noise power N is 10⫺11 W and the NEP can be decreased to less than 10⫺9 W兾冑Hz because of greatly reduced noise at such modulation frequencies [5,8] (see Fig. 9). At 250 GHz, as seen in Figs. 3 and 4, the GDD detected signal is approximately 20 times less than at 100 GHz. However, the EM wave power at 250 GHz is approximately an order of magnitude less than that at 100 GHz. This means that the NEP at 250 GHz is only slightly larger than that at 100 GHz. Other THz detectors, such as Golay cells and pyroelectric detectors, have a similar NEP. Efforts to consider the feasibility of developing focal plane arrays of GDD devices are under way. 3. Conclusions
The experimental results of this paper show that a GDD type N523 (green lamp) can serve as a THz detector. The low cost of the GDD, wide bandwidth, electronic ruggedness, and room-temperature operation can make it a preferred detector for THz applications in general and especially in THz imaging using a focal plane array of many inexpensive detectors with image sampling input to a computer for image display. The detection mechanism appears to be the THz radiation enhanced diffusion current (negative current) instead of the enhanced ionization current (positive current). This research was supported by the Office of Naval Research, Arlington, Virginia, USA, and by the U.S. Army Night Vision & Electronic Sensors Directorate, Ft. Belvoir, Virginia, USA, for which the authors are very grateful.
References 1. A. D. MacDonald, Microwave Breakdown in Gases (Wiley, 1966). 2. N. H. Farhat, “A plasma microwave power density detector,” Proc. IEEE 52, 1053–1054 (1964). 3. N. S. Kopeika, “Glow discharge detection of long wavelength electromagnetic radiation: cascade ionization process internal signal gain and temporal and spectral response properties,” IEEE Trans. Plasma Sci. PS-6, 139 –157 (1978). 4. D. C. McCain, “A plasma video detector,” IEEE Trans. Microwave Theory Tech. MMT-18, 66 – 67 (1970). 5. N. S. Kopeika and N. H. Farhat, “Video detection of millimeter waves with glow discharge tubes: Part I—Physical description; Part II—Experimental results,” IEEE Trans. Electron Devices ED-22, 534 –548 (1975). 6. N. S. Kopeika, “on the mechanism of glow discharge detection of microwave and millimeter wave radiation,” Proc. IEEE 63, 981–982 (1975). 7. N. H. Farhat and N. S. Kopeika, “A low-cost millimeter-wave glow-discharge detector,” Proc. IEEE 60, 759 –760 (1972). 8. N. S. Kopeika, “Noise spectra of commercial indicator glow discharge detectors,” Int. J. Electron. 39, 209 –218 (1975). 9. T. W. Crowe, J. L. Hesler, R. M. Weikle, and S. H. Jones, “GaAs devices and circuits for terahertz applications,” Infrared Phys. Technol. 40, 175–189 (1999). 10. G. Kozlov and A. Volkov, Millimeter and Submillimeter Wave Spectroscopy of Solids, Vol. 47 of Topics in Applied Physics, G. Gruner, ed. (Springer-Verlag, 1998). 11. J. A. Murphy and R. Padman, “Phase centers of horn antennas using Gaussian mode analysis,” IEEE Trans. Antennas Propag. 38, 1306 –1310 (1990). 12. P. F. Goldsmith, “Quasi optical technique at millimeter and sub-millimeter wavelengths,” in Infrared and Millimeter Waves, K. Button, ed. (Academic, 1982), Vol. 6, Chap. 5, pp. 277–343. 13. N. S. kopeika, “The influence of external electrodes in millimeter wave video detection with glow discharge plasmas,” in 1974 IEEE National Telecommunications Conference (IEEE, 1974), pp. 1069 –1073.
10 October 2007 兾 Vol. 46, No. 29 兾 APPLIED OPTICS
7211
