Open Access
Add to BibSonomyAbstract
We present a compact sensor head for a multi-channel terahertz (THz) spectroscopy system. A THz pulse generated by a photoconductive antenna is split into spatially separated sub-pulses, which have different transit times. The time-dependent order of the sub-pulses can be translated into a spatial resolution. By using only one pair of antennas the developed sensor head provides up to 20 individual measurement zones with full amplitude and phase information. The sensor head can be integrated into two boxes with a small footprint so that the system is well suited for industrial applications.
© 2016 Optical Society of America
1. Introduction
In comparison to other non-destructive testing procedures like metal detectors [1], ultrasonic sensors [2], x-rays or spectrometers using neighboring frequency ranges [3] THz time-domain spectroscopy (TDS) combines several unique characteristics [4]. For example, THz waves have a similar penetration depth in dielectric materials like microwaves but offer a high spatial resolution in the sub-millimeter range like infrared radiation. Therefore, THz TDS is widely discussed as a technique for monitoring of industrial production processes [5–11]. It has been shown, for instance, that THz TDS systems are suitable for monitoring the thickness of paper or paint during paper production [12, 13] or for detecting impurities in chocolate [14]. A large obstacle to put this technique into practice is the need for high-speed measurements, which cover a large area. This is required when implementing a quality control in many industrial production processes.
Classical THz TDS systems are based on a point source and a point detector. Hence, they only allow for one measurement point per detector/emitter antenna pair. With a typical measurement speed of several seconds for a single point, recording an image or measuring several points of a large sample can take several hours. Even if the measurement speed is increased to several 100 Hz [15–17], a measurement via a raster scanning approach with one antenna pair is too slow for many industrial applications.
There are several approaches to overcome this obstacle. A direct attempt to get a practical THz imaging system is the use of multiple emitter/detector antenna pairs [18]. Considering the cost of fiber coupled antenna modules, this approach is not cost-efficient. Besides methods for two-dimensional spatiotemporal THz imaging based on a regenerative amplifier have been demonstrated [19, 20]. Yet, they are more complex and expensive in comparison to the approach presented here. Further attempts include the use of oscillating mirrors combined with F-Theta-Lens-Systems [21, 22]. Yet, this approach is elaborate and an entire line-by-line scanning is not possible in this case. Also compressed sensing solutions for THz imaging [23, 24] are limited in their resolution since no fast spatial light modulators exists for the THz frequency range. Another THz imaging solution where the broadband frequency information of the THz pulses is transformed into a spatial resolution using a blazed diffraction grating [25] is limited by the absence of phase information.
Here, we present a concept, which allows for ten individual channels i.e. measurement zones with a nearly homogenous power distribution driven by only one pair of antennas. Each of these measurement zones provide a full amplitude and phase information of the sample. In principle, this scheme can be extended to twenty or more channels.
2. Concept of the sensor head
To realize the individual measurement zones with a nearly homogenous power distribution by using only one pair of antennas the developed concept is based on the partial reflection at a dielectric surface. For the dielectric material high-density polyethylene (HDPE) is chosen. HDPE has a low absorbance and a flat refractive index in the THz frequency range [26]. Hence, it is ideally suitable for our application.
The amount of the reflected radiation at an interface can be described by the well-known Fresnel equations for dielectrics. For s- and p-polarized radiation, the reflection coefficient can be described as follows (Eq. (1)):
Here, n1 is the refractive index of the material where the beam travels first and n2 is the refractive index of the material at whose surface the beam is reflected and refracted (c.f. inset Fig. 1). Based on Snell`s law, the angle β can be described in terms of α, n1 and n2.
Fig. 1 Reflection coefficient for an air/HDPE interface (red) and a HDPE/air interface (blue) for s-polarized radiation (solid lines) and for p-polarized radiation (dashed lines), inset: refraction and reflection at an interface between two media.
Figure 1 shows the reflection coefficient as a function of incidence angle for s- and p-polarization. It shall illustrate two for the sensor head important features of the reflection coefficient at an air (n = 1) /HDPE (n = 1.54) (red) interface and a HDPE/air interface (blue), respectively. First, for both geometries the reflection coefficient is nearly constant for incident angles smaller than 10°. Second, for these small angles there is only a small difference between the reflection at the HDPE/air interface and the air/HDPE interface. This is important, since both interfaces are used and a homogenous power distribution is desired.
The sensor head is composed of a set of HDPE beam splitters (c.f. Figure 2), which are designed such that each beam splitter reflects a well-defined amount of THz photons at each interface. The remaining fraction of the pulse propagates further to the adjacent beam splitter. Air/HDPE interfaces and HDPE/air interfaces are alternating. To ensure a nearly homogenous output power of the reflected sub-pulses the angles of incidence between the incoming beam and the beam splitters have to be in the range below 10 degrees, i.e. in the range where the reflection coefficient does not vary much with the incidence angle (c.f. Figure 1).
In this way up to 20 individual measurements channels can be obtained by using only one antenna pair. These channels cover 20 zones on the paper web. Hence, the technique suggested here represents a major progress towards a real-time 100% quality control in the production process.
3. Implementation of the concept
The most direct implementation of the concept is the use of uniform beam splitters with parallel surfaces. However, this entails two problems: first, the power of the different sub-pulses varies over a large range since the incident angle and the intensity of the incoming radiation change for each interface due to the refraction and reflection at the previous interface. The refraction poses also the second problem: the direction of the reflected sub-pulses is different for each interface. Overlaps between the different sub-pulses occur or if the distance between two beam splitters is not large enough a sub-pulse hits a previous beam splitter (c.f. Figure 3). Furthermore, unwanted back reflections occur for the reflections at HDPE/air interfaces so that the intensity of the sub-pulses decreases. As a consequence, each beam splitter has to be designed and optimized individually.
By using a numerical optimization based on a usual hill climbing algorithm the beam splitters can be designed in such a way that the power distribution of the sub-pulses is perfectly homogenous. But the optimization of the angles between the incoming beam and the beam splitters only in respect to the power distribution leads also again to large differences in the direction of the reflected sub-pulses. In our case a defined direction of the emerging beams is more important than a perfect homogenous power distribution. Thus, the direction of the sub-pulses is a criteria of exclusion in the optimization process so that a simple forward propagation simulation of the beam splitting device is more suitable in this case.
In Fig. 5 the results of the forward propagation simulation are shown. Provided that the direction of the sub-pulses is parallel (Fig. 4(a) blue dots) the tilt angles of the beam splitters are determined in respect to the angle of the refracted beam (Fig. 4(a) red squares) and for a nearly homogenous power distribution using an iteration loop (Fig. 4(a) blue triangles). The simulation result reveals that the output powers of the sub-pulses are in the range between 5.0 and 2.2% of the power of the initial pulse (c.f. Figure 4(b)) for s-polarized radiation. This variance is fully acceptable for the desired application and can be corrected after the measurement using a calibration function. For p-polarized radiation the reflected power is smaller and show larger differences between the two types of interfaces. Thus, the use of s-polarized THz radiation is preferred here. Additionally, the simulation of the optimized device has shown that the error of the reflected power that occurs if the incident angle changes by 3° is in the range of less than 0.01% for s-polarized radiation. Thus, the developed system is stable against differences in the incident angle.
Fig. 4 Simulation results of a system with 20 sub-pulses representing 20 channels (a) Angles for transmitted and reflected beam and for the beam splitter surface (b) Reflected power for s- and p-polarized light and deviation in the reflected power if the incident angle changes by 3° in case of s-polarized light.
Since the direction of the sub-pulses is not perpendicular to the sensor head the sub-pulses need to be redirected to ensure a vertical alignment of the sensor to the sample. Therefore, mirrors are added, which redirect the sub-pulses into a vertical plane (c.f. Figure 5).
Additionally, the individual sub-pulses emerging from the beam splitters show a non-linear time delay due to the different propagation velocity in HDPE and air. Hence, short and long time intervals between the sub-pulses alternate. Figure 6(a) illustrates this time-dependent order of the sub-pulses (grey) without any runtime adjustment.
Fig. 6 (a) Measured temporal positions of the sub-pulses before runtime adjustment (grey) and after (purple line) (b) Schematic of the runtime adjustment mirror unit.
To achieve an equidistant temporal spacing between the sub-pulses an additional time delay is required. Therefore, a second unit is added to the sensor head (c.f. Figures 6(b), 7(c) and 7(e)). This unit is composed of single mirrors in defined distances for each sub-pulse. To determine these distances a linear function is fitted to the time intervals. By calculating the difference between each time-position of the sub-pulses and the linear fit a corrective factor can be determined for the time position of each sub-pulse. This results in equidistant time delays between the different sub-pulses. The purple line in Fig. 6(a) illustrates the time intervals after runtime adjustment.
Fig. 7 (a) Top view of the emitter unit of the sensor head, (b) detailed view of a single beam splitter prototype, (c) lateral view of the runtime adjustment unit with focusing lens, (d) total view of the sensor head, (e) back view of the runtime adjustment unit with emitter unit above.
After the incoming THz radiation is split into sub-pulses, these sub-pulses have to be focused onto the detector antenna. Since the beams are distributed over a range of several decimeters this is achieved by a large lens with a thickness of 6 cm and a focal length of 1m. Yet, it was not possible to produce such a large lens without small aberrations. However, this is no problem for the application since only the reference and the sample pulse in one zone are compared to each other.
For industrial applications the system can be integrated into two separate units with a small footprint by folding the THz path. The emitter unit (c.f. Figures 5 and 7(a)) contains the photoconductive emitter antenna and the individual HDPE blocks for separating the pulses, while the receiver unit is composed of the mirror unit for the runtime adjustment of the spatially separated sub-pulses, the HDPE lens for collecting these sub-pulses and an integrated photoconductive detector antenna. The sample will be placed between these two units. Both units can be locked against dust so that they are well suited for an industrial environment.
4. Characterization
For the characterization of the sensor head the THz system has to fulfill two important requirement. Since the spatial distribution is encoded in the time delay, a large time window is required for the measurements with the sensor head. Therefore usual mechanical delay units are inappropriate due to their small time window of about 1000 ps. An asynchronous optical sampling THz system (ASOPS) provides a larger time window of up to 4 ns [27, 28]. In this type of spectrometer two femtosecond lasers are detuned slightly in their repetition rates resulting in a difference frequency and a time-dependent detector signal. The size of the time window depends on the repetition rate of the ASOPS system. Further details can be found in [27, 28]. The second requirement is the measurement speed. For production lines, a high measurement speed is called for. Also this requirement is fulfilled when an ASOPS THz system is used. Since there is no need for a mechanical delay line in an ASOPS THz system, faster measurements and a larger time window are possible in comparison to THz systems with mechanical delay lines. Here, we use a commercial ASOPS system from Menlo system working at the telecom wavelength together with fiber-coupled antenna modules [29].
The repetition rate of the two laser is adjusted such that they differ by 20 Hz. This results in a scanning range of 4 ns. Since the sub-pulses have a time spacing of approximately 159 ps, this scanning range would in principle allow for scanning 25 individual measurement zones simultaneously. For a first characterization the sensor head is composed of five beam splitter resulting in ten measurement zones. An extension to 20 measurement zones or more can be done later.
In a first step a raster scan image 10 cm in front of the runtime adjustment unit without a focusing lens was performed to verify that all sub-pulses are available. Therefore, the detector antenna performs a raster scan in front of the runtime adjustment unit. The measurement results in Fig. 8 show the peak-to-peak amplitude of the ten sub-pulses. The power of the sub-pulses decreases slightly with increasing number like expected (c.f. Figure 4(b)). Additionally, sub-pulse number six exhibits a smaller amplitude than expected. This is due to aberrations of the sensor head, since fixed mirrors are used to reduce the degrees of freedom. The sub-pulses are located at different positions in the image, since the sub-pulses are reflected at different positions at the runtime adjustment unit. By calculating the FWHM of the peak-to-peak amplitudes, we estimate a coverage of 82% for the scanned area. This measurement proves that the concept works as intended.
After this first characterization, the next step is to focus the sub-pulses onto the detector antenna. Therefore, the designed HDPE lens is added to the sensor head and a second measurement is performed. Here the detector antenna stays at the focal point of the lens. By using only one pair of antennas all ten sub-pulses are measured simultaneously. For this proof of concept measurement we averaged 1000 waveforms, i.e. much more than for a later in-line measurement. Hence, the data acquisition time was 50 s in this case. Furthermore, a bandpass filter was used to remove noise with higher frequencies.
Figure 9 illustrates the results of this second measurement. Here the spatial distribution of the sub-pulses is encoded in equidistant time intervals of 160 ps, which is in a good agreement with the calculated time intervals. The amplitude of the first and last sub-pulse is attenuated in comparison to the other sub-pulses. We attribute this to aberrations of the HDPE focusing lens, since this feature cannot be seen in Fig. 8. These aberrations of the focusing lens as well as the aberration of the sensor head can be corrected using a calibration function. For most industrial applications the signal-to-noise ratio is sufficient to detect different impurities in paper or other thin materials. Altogether, the proof of concept could be provided successfully.
5. Conclusion and outlook
We designed an innovative compact sensor head for a multichannel THz measuring system. A THz pulse generated by a photoconductive antenna is split into spatially separated sub-pulses, which have different equidistant transit times. The time-dependent order of the sub-pulses gives a spatial resolution by using only one pair of antennas. Instead of measuring the sample point by point like raster scanning approaches do this new sensor head covers a line of 30 cm simultaneously by concurrently low costs in comparison to antenna array solutions or solutions with regenerative amplifiers. Furthermore, a complete amplitude- and phase-information is provided for all individual measurement zones. Here, we demonstrated the concept for ten sub-pulses but the system can be extended to 20 or more sub-pulses. The zones can be arranged beneath to each other so that a complete area is monitored or they can be extended to a larger area depending on the application. In this way a 100% quality control is possible as well as controlling single parts of a sample. At the moment the data acquisition time is limited by the stability of the ASOPS system used and the transimpedance amplifier. However, ASOPS THz systems of the latest generation are more stable so that even faster data acquisition rates with a higher signal-to-noise ratio are possible. Therefore and due its compactness, the sensor head is well suited for industrial applications like contactless quality monitoring in production lines.
Funding
ZIM via the project WeDefIn (Project number: KF2376007AB4).
References and links
1. M. Graves, A. Smith, and B. Batchelor, “Approaches to foreign body detection in foods,” Trends Foods Sci. 9(1), 21–27 (1998). [CrossRef]
2. P. Hauptmann, N. Hoppe, and A. Püttmer, “Applications of ultrasonic sensors in the process industry,” Meas. Sci. Technol. 13(8), R73–R83 (2002). [CrossRef]
3. B. Batchelor, D. Hill, and H. Hodgson, Automated Visual Inspection (Elsevier Science Pub., 1985).
4. P. U. Jepsen, D. G. Cooke, and M. Koch, “Terahertz spectroscopy and imaging - modern techniques and applications,” Laser Photonics Rev. 5, 1–43 (2010).
5. S. Wietzker, C. Jördens, N. Krumbholz, B. Baudrit, M. Bastian, and M. Koch, “Terahertz imaging: a new non-destructive technique for the quality control of plastic weld joints,” J. Europ. Opt. Soc. 2, 07013 (2007). [CrossRef]
6. L.-L. Zhang, N. Karpowicz, C.-L. Zhang, Y.-J. Zhao, and X.-C. Zhang, “Real-time nondestructive imaging with THz waves,” Opt. Commun. 281(6), 1473–1475 (2008). [CrossRef]
7. D. Banerjee, W. von Spiegel, M. D. Thomson, S. Schabel, and H. G. Roskos, “Diagnosing water content in paper by terahertz radiation,” Opt. Express 16(12), 9060–9066 (2008). [CrossRef] [PubMed]
8. I. Duling and D. Zimdars, “Terahertz Imaging: Revealing hidden defects,” Nat. Photonics 3(11), 630–632 (2009). [CrossRef]
9. C. Jansen, S. Wietzke, O. Peters, M. Scheller, N. Vieweg, M. Salhi, N. Krumbholz, C. Jördens, T. Hochrein, and M. Koch, “Terahertz Imaging: Applications And Perspectives,” Appl. Opt. 49(19), E48–E57 (2010). [CrossRef] [PubMed]
10. A. Redo-Sanchez, N. Laman, B. Schulkin, and T. Tongue, “Review of Terahertz Technology Readiness Assessment and Applications,” J. Infrared Millim. Terahertz Waves 34(9), 500–518 (2013). [CrossRef]
11. T. Hochrein, “Markets, Availability, Notice, and Technical Performance of Terahertz Systems: Historic Development, Present, and Trends,” J. Infrared Millim. Terahertz Waves 36(3), 235–254 (2015). [CrossRef]
12. S. F. Busch, T. Probst, L. Duschek, R. Wilk, M. Voitsch, F. Fender, S. Lübbecke, G. Gärtner, V. P. Wallace, and M. Koch, “Inline Monitoring of Paper Thickness in an Industrial Setting,” in Infrared, Millimeter, and Terahertz Waves (IRMMW-THZ), 2013 38th International Conference on, 1–2 (2013). [CrossRef]
13. T. Yasui, T. Yasuda, K. Sawanaka, and T. Araki, “Terahertz paintmeter for noncontact monitoring of thickness and drying progress in paint film,” Appl. Opt. 44(32), 6849–6856 (2005). [CrossRef] [PubMed]
14. C. Jördens and M. Koch, “Detection of foreign bodies in chocolate with pulsed terahertz spectroscopy,” Opt. Eng. 47(3), 037003 (2008). [CrossRef]
15. T. Probst, A. Rehn, S. F. Busch, S. Chatterjee, M. Koch, and M. Scheller, “Cost-efficient delay generator for fast terahertz imaging,” Opt. Lett. 39(16), 4863–4866 (2014). [CrossRef] [PubMed]
16. D.-S. Yee, K. H. Jin, J. S. Yahng, H.-S. Yang, C. Y. Kim, and J. C. Ye, “High-speed terahertz reflection three-dimensional imaging using beam steering,” Opt. Express 23(4), 5027–5034 (2015). [CrossRef] [PubMed]
17. H. Guerboukha, A. Markov, H. Qu, and M. Skorobogatiy, “Time Resolved Dynamic Measurements at THz Frequencies Using a Rotary Optical Delay Line,” IEEE Trans. Terahertz Sci. Technol. 5(4), 564–572 (2015). [CrossRef]
18. B. Pradarutti, R. Müller, G. Matthäus, C. Brückner, S. Riehemann, G. Notni, S. Nolte, and A. Tünnermann, “Multichannel balanced electro-optic detection for Terahertz imaging,” Opt. Express 15(26), 17652–17660 (2007). [CrossRef] [PubMed]
19. T. Yasuda, T. Yasui, T. Araki, and E. Abraham, “Real-time two-dimensional terahertz tomography of moving objects,” Opt. Commun. 267(1), 128–136 (2006). [CrossRef]
20. T. Yasui, K. Sawanaka, A. Ihara, E. Abraham, M. Hashimoto, and T. Araki, “Real-time terahertz color scanner for moving objects,” Opt. Express 16(2), 1208–1221 (2008). [CrossRef] [PubMed]
21. D. Zimdars, G. Fichter, C. Megdanoff, J. Duquette, M. Murdock, I. Duling, J. White, and S. Williamson, “High-Speed Hand-Held Wide Aperture Time-Domain Terahertz Imaging System,” in CLEO Proceedings, p. CWO1 (2010). [CrossRef]
22. S. Katletz, M. Pfleger, H. Pühringer, N. Vieweg, B. Scherger, B. Heinen, M. Koch, and K. Wiesauer, “Efficient terahertz en-face imaging,” Opt. Express 19(23), 23042–23053 (2011). [CrossRef] [PubMed]
23. W. L. Chan, K. Charan, D. Takhar, K. F. Kelly, R. G. Baraniuk, and D. M. Mittleman, “A single-pixel terahertz imaging system based on compressed sensing,” Appl. Phys. Lett. 93(12), 121105 (2008). [CrossRef]
24. D. Shrekenhamer, C. M. Watts, and W. J. Padilla, “Terahertz single pixel imaging with an optically controlled dynamic spatial light modulator,” Opt. Express 21(10), 12507–12518 (2013). [CrossRef] [PubMed]
25. S. Schumann, C. Jansen, M. Schwerdtfeger, S. Busch, O. Peters, M. Scheller, and M. Koch, “Spectrum to space transformed fast terahertz imaging,” Opt. Express 20(17), 19200–19205 (2012). [CrossRef] [PubMed]
26. S. Sommer, T. Raidt, B. M. Fischer, F. Katzenberg, J. C. Tiller, and M. Koch, “THz-Spectroscopy on High Density Polyethylene with Different Crystallinity,” J. Infrared Millim. Terahertz Waves 37(2), 189–197 (2016). [CrossRef]
27. T. Yasui, E. Saneyoshi, and T. Araki, “Asynchronous optical sampling terahertz time-domain spectroscopy for ultrahigh spectral resolution and rapid data acquisition,” Appl. Phys. Lett. 87(6), 061101 (2005). [CrossRef]
28. A. Bartels, A. Thoma, C. Janke, T. Dekorsy, A. Dreyhaupt, S. Winnerl, and M. Helm, “High-resolution THz spectrometer with kHz scan rates,” Opt. Express 14(1), 430–437 (2006). [CrossRef] [PubMed]
29. R. J. B. Dietz, M. Gerhard, D. Stanze, M. Koch, B. Sartorius, and M. Schell, “THz generation at 1.55 µm excitation: six-fold increase in THz conversion efficiency by separated photoconductive and trapping regions,” Opt. Express 19(27), 25911–25917 (2011). [CrossRef] [PubMed]
