1. Introduction

Terahertz (THz) radiation detection has attracted extensive research interests and achieved great progress in recent years. Conventional thermal sensing elements, for instance, microbolometer and pyroelectric arrays, have become commercially available for direction THz detection [1]. However, sensitive, fast, room-temperature, low-cost, and easily integratable THz detectors are still highly desired for real-world applications of THz technology in the fields of sensing, security, biomedicine, and communications [2–4]. State-of-the-art detection technologies which are expected to meet these requirements focus on employing novel electronic or optoelectronic devices, such as complementary metal oxide semiconductor (CMOS) field-effect transistors (FETs) [5, 6] as well as carbon nanomaterial-based photodetectors [7–14]. The CMOS FETs are usually operated in the sub-THz range (<1 THz) limited by the intrinsic cut-off frequency, while the latter type of devices promise broadband photodetection of THz radiation.

Carbon nanomaterials, including carbon nanotubes (CNTs) and graphene, exhibit unique electrical and optical properties benefiting from their unusual crystal and electronic band structures [15]. The photon absorption ability of CNTs and graphene over a very broad spectral range makes them promising materials for the active elements in photodetectors from visible to THz frequencies [7–14,16–19]. Recently, THz-induced bolometric responses of metallic CNTs [7,8] at low temperatures and photothermoelectric (PTE) responses of heterojunction-like structures formed by asymmetric CNTs [10], CNT–metal contact [11], and p–n junction [12] at room temperatures have been reported, demonstrating the potential of using CNT-based devices for THz radiation detection. These devices either require antenna coupling for a small sensitive area or are made of macroscopic films. Different from the existing works, here we present a THz photodetector with significant photoresponse originating from the contact between a metal and a suspended, aligned, and macroscopically long double-walled CNT (DWCNT) bundle with a width on the wavelength scale. Due to the alignment of DWNTs and the centimeter length scale of the bundle, the photodetector exhibits strong polarization and position sensitive response, with a maximum responsivity about 10 times higher than our previous device fabricated with random network of DWCNTs [11]. We characterize the THz power dependence and temporal response of the detector and also analyze the results on the basis of PTE mechanism.

2. Experiments

The geometry of our device is illustrated in Fig. 1(a). It has a simple structure and consists of a U-shaped glass substrate, two metallic silver (Ag) electrodes and a suspended DWCNT macrobundle. First, we used a tweezer to pull out the nanotube bundle directly from a chemical vapor deposition-synthesized, large-scale DWCNT film [20] immersed in deionized water. Then we put the bundle over one arm of the U-shaped substrate and stretched it to another arm. Because the wet bundle was readily adsorbed onto the uppermost surface of the U-shaped substrate, it was naturally straightened to bridge the trench of the substrate and be kept in a suspended state. After the water evaporated, a layer of conductive Ag paint was deposited on the two ends of the bundle. Thus the DWCNT–Ag heterojunctions were formed at the contacts between the bundle and the metal electrodes. The macrobundle has a length of more than 1 cm and a nearly uniform width of about 65 μm, as seen in the optical and scanning electron microscopy (SEM) images [Figs. 1(b) and 1(c)], where thousands of nanotubes are aligned along the longitudinal direction [21]. SEM images of the DWCNT–Ag contacts reveal that the nanotubes slightly overlap with the electrode at the junction region, as shown in Figs. 1(d) and 1(e). A far-infrared gas laser (FIRL 100, Edinburgh Instruments Ltd.) served as the illumination source, which can produce continuous-wave THz radiation with many discrete wavelengths. Here we used the methanol lines at 163 μm (1.84 THz), 118.8 μm (2.52 THz), 96.5 μm (3.11 THz), and 70.6 μm (4.25 THz), with maximum powers of about 20 mW, 90 mW, 150 mW, and 35 mW, respectively. The THz beam was focused onto the detector with a focal spot diameter of approximately 1.5 mm (full width at half maximum, FWHM), as measured by the scanning knife-edge method for 2.52 THz [see Fig. 1(f)]. The total incident THz power was monitored by a calibrated pyroelectric detector (Sensor- und Lasertechnik, THz20). We used a sourcemeter (Keithley 2602B) to measure the photocurrent generated in the device. The current is positive when it flows in the DWCNT bundle from right (junction B) to left (junction A). Therefore, the carrier transport direction in the sample could be determined according to the sign of the current. All the measurements were performed in ambient air at room temperature. In the experiments except for that of evaluating the position effect, the THz light spot was located at the contact region displaying the maximum photocurrent.

 

Fig. 1 (a) Schematic diagram of the device geometry and the experimental configuration. (b) Optical image of the device consisting of a DWCNT macrobundle contacted with two Ag electrodes and suspended over a 10 mm trench. (c) SEM image of the DWCNT bundle with a width of 65 μm. (d) and (e) SEM images of the DWCNT–Ag contacts for (d) junction A, and (e) junction B. (f) Measurement of the THz beam spot size; the solid curve is a Gaussian fit of the experimental data.

Download Full Size | PDF

3. Results and discussion

First, we tested the current–voltage (I − V) characteristics of our device. The black line of Fig. 2(a) represents a typical dark I − V relationship without THz illumination, where its nearly linear behavior over a larger bias voltage range (from −1 V to 1 V, not shown here) implies that the nanotube bundle has an ohmic contact with the electrode. The total resistance of the device calculated from this measurement is about 12.3 Ω. Such relatively small resistance is not surprising considering the metallic character of DWCNTs [21]. Upon exposure to 2.52 THz radiation with an actual received power of about 4 mW (see below) for junctions A and B, the I − V curves shift downward and upward, respectively, as shown by the red and blue lines of Fig. 2(a). Obviously, a short-circuit photocurrent (I_SC) at zero bias as well as an open-circuit photovoltage (V_OC) at zero current are produced under THz excitation. We determine that I_SC is ∼65 μA and V_OC is ∼800 μV for junction B. In the presence of THz radiation, there is a slight change (∼1.6%) in the slope of the I − V curve, indicating that the bolometric effect does not dominate the THz photoresponse and could be ignored here [12]. In addition, we find that the photocurrent directions are opposite for the two junctions. These observations are consistent with the case of visible and infrared light illumination for both film and bundle samples [17, 18, 21, 22].

 

Fig. 2 (a) I − V characteristics of the detector in the dark (black line) and under THz illumination on junction A (red line) and junction B (blue line) with 4-mW power at 2.52 THz. (b) Polarization dependence of the normalized photocurrent for illuminating junction B at 2.52 THz (0° corresponds to the polarization parallel to the nanotubes); the solid curve is a cosine fit of the experimental data. (c) Power dependence of the photocurrent for illuminating junction B at frequencies of 1.84 THz, 2.52 THz, 3.11 THz, and 4.25 THz, respectively; the solid and dashed lines are linear fit of the data at 2.52 THz and 3.11 THz. Inset: photocurrent responsivities (R_I) for the four frequencies.

Download Full Size | PDF

Benefiting from the alignment of DWCNTs in the macrobundle, the THz response of our device exhibits strong dependence on the polarization of the incident beam, as demonstrated in Fig. 2(b), where the photocurrent I_SC generated by illuminating junction B at 2.52 THz versus the angle between the THz polarization and the nanotube alignment direction is plotted. This phenomenon arises from the polarization-dependent absorption of CNTs [12,23]. I_SC is normalized to its maximum intensity which corresponds to the polarization parallel to the nanotubes (i.e., 0°). The photocurrent ratio between parallel and perpendicular polarizations is around 1.4, not high due to the imperfect alignment of nanotubes in our sample and also the fact that in the perpendicular direction the induced currents might be not negligible. Moreover, we evaluated the dependence of I_SC on the incident THz power. Figure 2(c) shows I_SC as a function of the power actually received by nanotubes for four frequencies (1.84 THz, 2.52 THz, 3.11 THz, and 4.25 THz), revealing the broadband nature of the detector. The power was estimated by integrating the intensity distribution of the THz beam spot over the area occupied by the DWCNT bundle. The THz spot intensity was assumed to have a Gaussian distribution with a FWHM obtained from Fig. 1(f) and a total power recorded by the THz20 detector. The differences in spot sizes for the four different THz frequencies were ignored here. It should be pointed out that the electrode will also contribute to the actual received power of the junction, yielding an underestimate of the power, which is not considered in the present work. Good linearity of the response is observed in the measured power range. From the slope of the linearly fitted line, we can obtain the photocurrent responsivities, whose values are 13.5 mA/W, 16.2 mA/W, 17.9 mA/W, and 18.7 mA/W at 1.84 THz, 2.52 THz, 3.11 THz, and 4.25 THz, respectively, as given in the inset of Fig. 2(c).. The corresponding photovoltage responsivities are 0.17 V/W, 0.2 V/W, 0.22 V/W, and 0.23 V/W, respectively. These values are about 10 times higher than our previous device fabricated with substrate-bound random network of DWCNTs [11]. The improvement of the responsivity for a suspended geometry considered here coincides with previous studies on the effect of the thermal environment based on a PTE model [12, 17, 24]. Consequently, the PTE mechanism may be the origin of the THz-induced photoresponse of our device.

Furthermore, we tested the time-dependent response of the detector by periodically chopping the incident beam under ten different modulation frequencies. Figures 3(a) to 3(c) give the time evolution of the photosignal for modulation frequencies of 0.0083 Hz (120 s period), 2 Hz, and 20 Hz. The photosignal waveforms for other frequencies under test are included in the Appendix. It is clear that as the THz light on/off rate increases, the absolute intensity and relative variation of the photocurrent decrease with an expected exponential relation between the peak-to-peak amplitude of the photocurrent and the modulation frequency, as shown in Fig. 3(d). Fitting the first rising edge of the curve in Fig. 3(a) with a summation of two exponential functions, we obtain two time constants, 0.495 s and 13.74 s, which are larger than that of the non-suspended sample [11]. This is also consistent with a previous observation that a smaller substrate thermal conductivity would lead to a slower temporal PTE response [17, 24]. In another aspect, the response time has been found to possess a quadratic dependence on the suspended length [25] and it is not surprising why the response is slow for the centimeter-long CNT bundle concerned here. Therefore, in terms of temporal analysis, the possibility of PTE effect dominating the response of our device is demonstrated again.

 

Fig. 3 Temporal photocurrent response of the detector for illuminating junction B at 2.52 THz under modulation frequencies of (a) 0.0083 Hz, (b) 2 Hz, and (c) 20 Hz. (d) Peak-to-peak amplitude of the photocurrent as a function of the modulation frequency; the solid line is a linear fit for the semilog plot of the data.

Download Full Size | PDF

In order to further confirm that the PTE nature is the main contribution, we performed scanning photocurrent experiments by mechanically moving the detector step by step along the longitudinal direction of the DWCNT bundle to change the THz beam position between the two electrodes. Figure 4 illustrates the position-dependent I_SC profiles of the device with its original geometry, as well as a modified one with the Ag electrodes being covered by two sheets of nickel (see inset) where there are no electrical contact and thermal transport between them. In the case of bare electrode, the photoresponse exhibits significant enhancement at the DWCNT–Ag contacts and opposite sign for the two junctions, as already mentioned in the I − V measurements. Such characteristics reproduce the general behaviors previously reported for macroscopic CNT films [17,18,24,25], CNT macrobundles [21,22], and a single suspended CNT [26, 27] under visible light illumination. The left and right peak amplitudes are not on the same level, which probably arises from the different contact qualities of the junctions. When the THz spot is located away from the electrode edge and not incident on the suspended nanotubes, a response of I_SC on the order of several μA still exists, excluding the possibility of a photovoltaic effect [26]. These observations agree well with the PTE mechanism discussed elsewhere [17, 18, 24–26]. If the electrode is covered and only the CNT bundle could be illuminated, the I_SC intensity around the contact is roughly reduced to a half and the positions of its maxima become away from the electrode edges, indicating the importance of the metal electrode in photocurrent generation. Note that the electrodes do not only play a role in the PTE process, but also may serve as antennas and thereby increase the coupling efficiency of the incident radiation.

 

Fig. 4 Scanning photocurrent measurements of the device with its original geometry (blue curve) and a modified one (red curve). Dashed lines show the edges of the electrodes. Inset: schematic diagram of the device [side view, see Fig. 1(a)] modified by covering the Ag electrodes with two sheets of nickel and ensuring no electrical contact and thermal transport between them.

Download Full Size | PDF

According to the PTE theory, the rise and decay of the position-dependent I_SC profile can be described by a parameter named thermal length scale λ [17, 18]. For substrate-bound CNTs, λ is found to have a specific relation to the substrate thermal conductivity κ_sub, that is $\lambda \propto {\kappa }_{\text{sub}}^{-1/2}$. By using a proportional factor of 3×10⁻⁴ m^1/2W^1/2/K^1/2 extracted from Ref. 16 and assuming air as the substrate with κ_air = 0.025 W/m·K for suspended CNTs, we estimate that λ is about 2 mm. This value will increase after accounting for the THz beam spot size, and thus is close to that obtained from Fig. 4. The red curve in Fig. 4 shows faster decay in the electrode region due to no thermal contribution from it and only the impact of THz spot size.

The origin of the photocurrent or the photovoltage generated by the PTE effect stems from light-induced heating of an interface junction between two materials with different Seebeck coefficients. Here the junction corresponds to the DWCNT–Ag contact. Considering the case of THz beam illuminating junction B, the photovoltage (total thermoelectric potential difference between the two electrodes) can be described by a simple model as follows [18, 24]:

It is evident that all the above measurements and analysis support the PTE mechanism of the detector photoresponse. Indeed, the responsivity of our detector is lower than previous results on a CNT-based p–n junction detector [12], since a larger Seebeck coefficient difference exists between the p-type and n-type CNTs [17]. We expect that improvement of the sensitivity would be achieved by optimizing the thermoelectric properties of DWCNTs, the structure and material of the metal electrode, and the contact quality of the heterojunction.

4. Conclusion

In conclusion, we have developed a room-temperature, polarization-sensitive, and broadband THz photodetector realized with a heterojunction formed by suspended DWCNT macrobundle and metal contact. The THz-induced photoresponse was found to result from the polarization dependent absorption of THz radiation by aligned DWCNTs in a broad spectral range and subsequent heating of the junction region. A significant photocurrent/photovoltage was then generated through the PTE effect which was confirmed by scanning photocurrent measurements in combination with characterization of the sensitivity and temporal response of the device. The proposed detector has a very simple structure and the wavelength-scale lateral size of the DWCNT bundle ensures its high compactness. Our work opens a new way of using CNT-based optoelectronic devices for THz detection.

5. Appendix

In order to test the temporal photocurrent response of the THz photodetector, we periodically chopped the incident THz beam under different modulation frequencies. Figure 5 shows the waveforms of the photocurrent for modulation frequencies of 0.05 Hz, 0.1 Hz, 5 Hz, 10 Hz, 15 Hz, 25 Hz, and 30 Hz. From these results, we can extract the peak-to-peak amplitude of the photocurrent signal and quantitatively evaluate the temporal response of the detector.

 

Fig. 5 Temporal waveforms of the photocurrent under modulation frequencies of (a) 0.05 Hz, (b) 0.1 Hz, (c) 5 Hz, (d) 10 Hz, (e) 15 Hz, (f) 25 Hz, and (g) 30 Hz.

Download Full Size | PDF

Acknowledgments

This work was supported by the National Natural Science Foundation of China (Grant Nos. 11404182, 11174172, and 11174170) and the National Science and Technology Support Program of China (Grant No. 2013BAK14B03).

References and links

1. A. Rogalski and F. Sizov, “Terahertz detectors and focal plane arrays,” Opto-Electron. Rev. 19, 346–404 (2011). [CrossRef]  

2. J. F. Federici, B. Schulkin, F. Huang, D. Gary, R. Barat, F. Oliveira, and D. Zimdars, “THz imaging and sensing for security applications—explosives, weapons and drugs,” Semicond. Sci. Technol. 20, S266–S280 (2005). [CrossRef]  

3. Z. D. Taylor, R. S. Singh, D. B. Bennett, P. Tewari, C. P. Kealey, N. Bajwa, M. O. Culjat, A. Stojadinovic, H. Lee, J.-P. Hubschman, E. R. Brown, and W. S. Grundfest, “THz medical imaging: in vivo hydration sensing,” IEEE Trans. THz Sci. Technol. 1, 201–219 (2011). [CrossRef]  

4. J. Federici and L. Moeller, “Review of terahertz and subterahertz wireless communications,” J. Appl. Phys. 107, 111101 (2010). [CrossRef]  

5. R. Al Hadi, H. Sherry, J. Grzyb, Y. Zhao, W. Förster, H. M. Keller, A. Cathelin, A. Kaiser, and U. R. Pfeiffer, “A 1 k-pixel video camera for 0.7–1.1 terahertz imaging applications in 65-nm CMOS,” IEEE J. Solid-State Circuit 47, 2999–3012 (2012). [CrossRef]  

6. S. Boppel, A. Lisauskas, M. Mundt, D. Seliuta, L. Minkevičius, I. Kašalynas, G. Valušis, M. Mittendorff, S. Winnerl, V. Krozer, and H. G. Roskos, “CMOS integrated antenna-coupled field-effect transistors for the detection of radiation from 0.2 to 4.3 THz,” IEEE Trans. Microw. Theory Tech. 60, 3834–3843 (2012). [CrossRef]  

7. K. Fu, R. Zannoni, C. Chan, S. H. Adams, J. Nicholson, E. Polizzi, and K. S. Yngvesson, “Terahertz detection in single wall carbon nanotubes,” Appl. Phys. Lett. 92, 033105 (2008). [CrossRef]  

8. J. D. Chudow, D. F. Santavicca, C. B. McKitterick, D. E. Prober, and P. Kim, “Terahertz detection mechanism and contact capacitance of individual metallic single-walled carbon nanotubes,” Appl. Phys. Lett. 100, 163503 (2012). [CrossRef]  

9. L. Vicarelli, M. S. Vitiello, D. Coquillat, A. Lombardo, A. C. Ferrari, W. Knap, M. Polini, V. Pellegrini, and A. Tredicucci, “Graphene field-effect transistors as room-temperature terahertz detectors,” Nat. Mater. 11, 865–871 (2012). [CrossRef]   [PubMed]  

10. G. Fedorov, A. Kardakova, I. Gayduchenko, I. Charayev, B. M. Voronov, M. Finkel, T. M. Klapwijk, S. Morozov, M. Presniakov, I. Bobrinetskiy, R. Ibragimov, and G. Goltsman, “Photothermoelectric response in asymmetric carbon nanotube devices exposed to sub-terahertz radiation,” Appl. Phys. Lett. 103, 181121 (2013). [CrossRef]  

11. Y. Wang, G. Zhang, L. Qiao, J. Wei, J.-L. Zhu, Z. Chen, Z. Zhao, and J.-L. Sun, “Photocurrent response of carbon nanotube–metal heterojunctions in the terahertz range,” Opt. Express 22, 5895–5903 (2014). [CrossRef]   [PubMed]  

12. X. He, N. Fujimura, J. M. Lloyd, K. J. Erickson, A. A. Talin, Q. Zhang, W. Gao, Q. Jiang, Y. Kawano, R. H. Hauge, F. Léonard, and J. Kono, “Carbon nanotube terahertz detector,” Nano Lett. 14, 3953–3958 (2014). [CrossRef]   [PubMed]  

13. M. Mittendorff, S. Winnerl, J. Kamann, J. Eroms, D. Weiss, H. Schneider, and M. Helm, “Ultrafast graphene-based broadband THz detector,” Appl. Phys. Lett. 103, 021113 (2013). [CrossRef]  

14. X. Cai X, A.B. Sushkov, R. J. Suess, M. M. Jadidi, G. S. Jenkins, L. O. Nyakiti, R. L. Myers-Ward, S. Li, J. Yan, D. K. Gaskill, T. E. Murphy, H. D. Drew, and M. S. Fuhrer, “Sensitive room-temperature terahertz detection via the photothermoelectric effect in graphene,” Nat. Nanotechnol. 9, 814–819 (2014). [CrossRef]   [PubMed]  

15. R. R. Hartmann, J. Kono, and M. E. Portnoi, “Terahertz science and technology of carbon nanomaterials,” Nanotechnology 25, 322001 (2014). [CrossRef]   [PubMed]  

16. F. Xia, T. Mueller, Y. Lin, Y. Lin, A. Valdes-Garcia, and P. Avouris, “Ultrafast graphene photodetector,” Nat. Nanotechnol. 4, 839–843 (2009). [CrossRef]   [PubMed]  

17. X. He, X. Wang, S. Nanot, K. Cong, Q. Jiang, A. A. Kane, J. E. M. Goldsmith, R. H. Hauge, F. Léonard, and J. Kono, “Photothermoelectric p–n junction photodetector with intrinsic broadband polarimetry based on macroscopic carbon nanotube films,” ACS Nano 7, 7271–7277 (2013). [CrossRef]   [PubMed]  

18. S. Nanot, A. W. Cummings, C. L. Pint, A. Ikeuchi, T. Akiho, K. Sueoka, R. H. Hauge, F. Léonard, and J. Kono. Broadband, “Polarization-sensitive photodetector based on optically-thick films of macroscopically long, dense, and aligned carbon nanotubes,” Sci. Rep. 3, 1335 (2013). [CrossRef]  

19. Y. Cao, J. Zhu, J. Xu, J. He, J.-L. Sun, Y. Wang, and Z. Zhao, “Ultra-broadband photodetector for the visible to terahertz range by self-assembling reduced graphene oxide-silicon nanowire array heterojunctions,” Small 10, 2345–2351 (2014). [CrossRef]   [PubMed]  

20. J. Wei, H. Zhu, Y. Li, B. Chen, Y. Jia, K. Wang, Z. Wang, W. Liu, J. Luo, M. Zheng, D. Wu, Y. Zhu, and B. Wei, “Ultrathin single-layered membranes from double-walled carbon nanotubes,” Adv. Mater. 18, 1695–1700 (2006). [CrossRef]  

21. J. Wei, J.-L. Sun, J.-L. Zhu, K. Wang, Z. Wang, J. Luo, D. Wu, and A. Cao, “Carbon nanotube macrobundles for light sensing,” Small 2, 988–993 (2006). [CrossRef]   [PubMed]  

22. J.-L. Sun, J. Wei, J.-L. Zhu, D. Xu, X. Liu, H. Sun, D.-H. Wu, and N.-L. Wu, “Photoinduced currents in carbon nanotube/metal heterojunctions,” Appl. Phys. Lett. 88, 131107 (2006). [CrossRef]  

23. L. Ren, C. L. Pint, L. G. Booshehri, W. D. Rice, X. Wang, D. J. Hilton, K. Takeya, I. Kawayama, M. Tonouchi, R. H. Hauge, and J. Kono, “Carbon nanotube terahertz polarizer,” Nano Lett. 9, 2610–2613 (2009). [CrossRef]   [PubMed]  

24. B. C. St-Antoine, D. Ménard, and R. Martel, “Photothermoelectric effects in single-walled carbon nanotube films: reinterpreting scanning photocurrent experiments,” Nano Res. 5, 73–81 (2012). [CrossRef]  

25. B. C. St-Antoine, D. Ménard, and R. Martel, “Position sensitive photothermoelectric effect in suspended single-walled carbon nanotube films,” Nano Lett. 9, 3503–3508 (2009). [CrossRef]   [PubMed]  

26. T. DeBorde, L. Aspitarte, T. Sharf, J. W. Kevek, and E. D. Minot, “Photothermoelectric effect in suspended semiconducting carbon nanotubes,” ACS Nano 8, 216–221 (2014). [CrossRef]  

27. M. Barkelid and V. Zwiller, “Photocurrent generation in semiconducting and metallic carbon nanotubes,” Nat. Photonics 8, 47–51 (2013). [CrossRef]  

28. T. Miao, W. Ma, X. Zhang, J. Wei, and J. Sun, “Significantly enhanced thermoelectric properties of ultralong double-walled carbon nanotube bundle,” Appl. Phys. Lett. 102, 053105 (2013). [CrossRef]  

29. S. Kasap, “Thermoelectric effects in metals: thermocouples,” An e-Booklet; University of Saskatchewan: Saskatoon, SK, 2001; http://kasap3.usask.ca/samples/Thermoelectric-Seebeck.pdf.

30. I. Maeng, C. Kang, S. J. Oh, J. H. Son, K. H. An, and Y. H. Lee, “Terahertz electrical and optical characteristics of double-walled carbon nanotubes and their comparison with single-walled carbon nanotubes,” Appl. Phys. Lett. 90, 051914 (2007). [CrossRef]