A high power, frequency-tunable THz source based on intracavity stimulated polariton scattering (SPS) in RbTiOPO4 (RTP) is demonstrated for the first time. Frequency tunable THz output was obtained from 3.10 to 4.15 THz, with a gap at 3.17 to 3.49 THz, arising from the 104 cm−1 A1 mode in RTP. A maximum average output power of 16.2 µW was detected at 3.8 THz. This is the highest average output power ever reported for an intracavity polariton laser.
© 2016 Optical Society of America
Spectroscopy and spectral imaging utilizing THz radiation have successfully demonstrated the potential to address a wide variety of revolutionary applications in industry, homeland security and medical sciences [1,2]. Human skin cancer detection , tumor detection and differentiation in mice , non-destructive testing for industrial quality control , concealed weapons detection  and toxic substances indentification  are among the numerous examples reported in the literature [1,8].
While the number of applications continue to build, their translation into mainstream use has been limited and this is due, in large part, to the lack of practical THz sources. Most of the available sources utilize a time-domain THz spectroscopy configuration (TDS-THz), in which a femtosecond laser pulse irradiates a photoconductive antenna, emitting a very short, broadband THz pulse (typically spanning several terahertz) . These TDS-THz systems rely on high power and expensive laser systems, as well as complex optomechanics and computer-controlled delay lines for detection. Development of reliable, robust and cost-effective THz sources with narrow band and frequency tunable output which can be interfaced with conventional THz detectors is then imperative, and the approach considered here uses a conventional diode-pumped Nd laser to produce such frequency-tunable THz radiation via stimulated polariton scattering (SPS) in nonlinear optical crystals .
Interaction of a fundamental laser (ωf) with a polariton mode of a crystal may generate, via SPS, Stokes (ωS) and THz (ωT) fields. Conservation of energy (ωf = ωS + ωT) and momentum (kf = kS + kT) in the process yields angle dependent frequencies for Stokes and THz radiation. As a result, variation in the interaction angle enables frequency-tunable THz output . Stimulated polariton scattering is a versatile technique to access the THz portion of the electromagnetic spectrum, and various pulsed systems using both extra-cavity and intra-cavity configurations have been developed [11–14]. Recently, a continuous-wave, frequency-tunable THz source based on intracavity SPS in lithium niobate (LiNbO3) crystals has also been reported , further demonstrating the flexibility of the technique for generating THz emission with different output modalities. The vast majority of SPS THz systems demonstrated to-date have been based on polariton modes in Mg:LiNbO3 [16–18], typically generating output between 0.6 to 3.2 THz. This tuning range is determined by the material properties of the Mg:LiNbO3, in particular the dispersion curve and absorption characteristics.
By employing different SPS-active crystals, different THz frequency ranges can be accessed, which may lead to enhanced uptake of these THz systems. In this context, extracavity SPS systems have been developed utilizing KTP (KTiOPO4) [19,20] and KTA (KTiOAsO4)  crystals. In KTP, three tunable bands were obtained when pumping with a 1064 nm, near-infrared laser: 3.17 to 3.44 THz, 4.19 to 5.19 THz and 5.55 to 6.13 THz. Also in KTP, four tunable bands were reported pumping using a laser at 532 nm: 5.7 to 6.1, 7.4 to 7.8, 11.5 to 11.8 and 13.3 to 13.5 THz. Similar THz emission was observed in KTA, with output tunable in five discrete bands between 3.59 to 6.43 THz. Discontinuities in emitted THz spectra are attributed to infrared absorption by A1 crystalline modes in the crystal. The efficiency of these KTP and KTA extracavity systems was broadly similar to extracavity systems based on Mg:LiNbO3 .
Rubidium titanyl phosphate (RbTiOPO4, RTP) is another SPS-active nonlinear crystal with high transparency range and higher damage threshold then KTP, making it ideal for high-power intracavity systems . An isomorph to KTP, RTP is orthorhombic at room temperature and belongs to C2υ (mm2) point group, with 47 A1 transverse optical modes both Raman and infrared active in the crystallographic z-axis direction. Polariton A1 modes are located at 104, 142.0, 159, 163, 211 and 269 cm−1 , and can be used for polariton scattering in RTP. Similar to what is observed in its isomorphs, the RTP crystal is expected to produce THz radiation in frequency tunable bands, with gaps located at the A1 modes below 269 cm−1.
With a desire to produce compact, efficient and diode-pumped THz sources that generate THz output beyond what is achieved with Mg:LiNbO3, in this work we report a frequency-tunable THz source based on intracavity SPS in an RTP crystal. This is the first time a nonlinear material different from lithium niobate has been used for intracavity SPS, and the maximum average THz power we report is in fact higher than has been previously reported for any intracavity SPS laser. This SPS laser produces THz output with tunability extending to 4.15 THz, well beyond what can be achieved in Mg:LiNbO3, in a simple and compact layout, and pumped with only 8 W from a laser diode. It represents a significant improvement towards broadening the spectral coverage of THz sources.
2. Experimental setup
The laser resonator is represented in Fig. 1. The fundamental field at 1064 nm was produced in a conventional diode end-pumped, Nd:YAG Q-switched laser resonator, containing an intracavity RTP crystal. The Stokes field was resonated in a separate laser cavity formed around the RTP. Wavelength tunability for Stokes and THz radiation was achieved by adjusting the angle θext between fundamental and Stokes resonator axes.
The fundamental resonator was end-pumped by a continuous-wave fiber coupled diode laser (100 µm core diameter, 0.22 NA) operated at maximum output power of 10 W at 808 nm. To reduce thermal lensing, the laser diode was chopped by an optical chopping wheel at 200 Hz and 50% duty cycle . Pump optics produced an approximate 300 µm diameter spot inside the 5 mm long, 5 mm diameter 1 a.t. % Nd:YAG crystal, which was anti-reflection (AR) coated for 808 and 1064 nm. The fundamental cavity was formed by flat mirror M1 (T > 99% at 808 nm and R > 99.99% at 1064nm) and concave (1000 mm ROC) mirror M2 having a high reflectivity (HR) coating (R = 99.4% at 1064 nm), separated by 230 mm. Pulsed operation was achieved with an acousto-optic Q-switch at 3 kHz repetition rate (a repetition rate much higher than diode laser chopping frequency to avoid any influence of the pump duty cycle on the Q-switching operation). An intracavity x-cut RTP crystal, 4x4x20 mm3 (Crystal Laser S. A.) with end faces AR coated for 1064-1100 nm (R < 0.1% at 1064 nm) was used as the SPS-crystal (fundamental laser field polarized parallel to the crystal z-axis).
The Stokes resonator was 85 mm in length, formed by two flat mirrors M3 and M4 coated HR from 1064 to 1090 nm (R>99.9%). Mirrors M3 and M4 were D-shaped to avoid clipping of the fundamental beam, and were mounted on a rotational stage, enabling fine tuning of the angle between Stokes and fundamental resonator axes (θext), promoting the consequent frequency tuning in Stokes and THz outputs . To avoid total internal reflection of the THz field inside the RTP, the crystal side face (y-surface) was polished and three high resistivity Silicon prisms (R > 10 kΩ/cm; 7 mm hypotenuse, 32°/58°/90° internal angles) were bonded .
Fundamental and Stokes output power leaking from M2 and M4, respectively, were monitored with a laser power meter (Thorlabs PM100D). Spectral content was detected with a high resolution fiber coupled grating spectrometer (Ocean Optics HR4000), and temporal profiles were measured with silicon photodiodes (Thorlabs DET10A/M; 1 ns rise time). A Golay cell (Tydex, GD-1P) in combination with a long pass filter (Tydex, LPF 14.3) and a 50 mm focal length collecting lens (Tydex, 50 mm EFL TPX lens) was used to measure the THz output from the Si prism array.
3.1 THz and Stokes fields frequency tunability
The external angle (θext) between fundamental and Stokes resonator axes was adjusted from ~1.8 to 6.3° (~1 to 3.6° internal angle), and a corresponding Stokes wavelength tunable from 1076.2 to 1091.7 nm was measured. From this, we calculated the corresponding THz frequencies to range from 3.10 to 7.03 THz. This is plotted in Fig. 2. This data was collected with the system operating with an incident diode pump power of 7 W; the fundamental laser wavelength was 1064.4 nm. Discontinuities in the Stokes wavelength tuning range were observed (Fig. 2), and these correlate with RTP infrared absorbing A1 modes at 104, 142, 159, 163 and 211 cm−1. Fine tuning of the external angle was performed close to these modes and the Stokes radiation spectra were collected. As this tuning was performed, a clear jump was observed at each mode, with the coexistence of dual-Stokes wavelengths either side of the gaps, as depicted in Fig. 3. This particular behavior suggests simultaneous, dual-terahertz output, which could be explored for applications involving differential transmission with THz radiation.
The THz frequency tuning curve is shown in Fig. 4, in which the THz frequency has been inferred from the Stokes wavelength. THz radiation was detected for Stokes wavelengths in the range of 1076.6 to 1080.3 nm, corresponding to a tunable range from 3.10 to 4.15 THz. The reported THz output power from the laser has been corrected for the frequency-dependent transmission loss of the filter and lens attached to the Golay cell, using data supplied by Tydex, and corresponds to the free space THz field after exiting the prisms. As expected, no THz emission was observed in the 104 cm−1 gap (THz frequencies between 3.16 THz and 3.5 THz). The significant dips in power observed around 3.68, 3.84 and 4.01 THz match water vapor absorption lines , and are attributed to the presence of water vapor in our laboratory environment. No efforts had been made taken to exclude water vapor. Somewhat surprisingly, no THz power was detected above 4.15 THz, the possible reasons for which are discussed later in this paper.
3.2 Laser power scaling
The most efficient THz emission was obtained at 3.80 THz, and power scaling measurements at this frequency are plotted in Fig. 5 for fundamental, Stokes and THz fields. The efficiency of the SPS process can be inferred from the fundamental field depletion, because it is related to the amount of fundamental power channeled into the SPS process. The fundamental field depletion was calculated by measuring the 1064 nm laser power with Stokes cavity enabled and disabled, as shown in Fig. 6. The Stokes output power versus fundamental field depletion is also shown as an inset in Fig. 6. The temporal profiles of the fundamental (depleted and undepleted, that is with and without SPS activity, respectively) and Stokes fields were measured when the system was pumped at 6 W, and are plotted in Fig. 7.
A maximum average power of 16.2 µW was obtained at 3.80 THz, with fundamental field depletion in excess of 60%. This is, to our knowledge, the highest average THz output power, and the highest fundamental field depletion ever reported for an intracavity SPS laser, suggesting very good mode matching between interacting fields. Within our data range, there is a near-linear relation between the Stokes output power and fundamental field depletion as can be observed in the inset of Fig. 6. At 6 W pump power, the fundamental field pulse width decreased from ~50 ns to ~21 ns (FWHM) when the SPS process was enabled, and a pulse width of ~17 ns was measured for the Stokes radiation.
4. Limiting factors on THz frequency tuning range
Tuning to higher THz frequencies resulted in a considerable reduction in detected THz output power, and no power could be detected for frequencies above 4.15 THz. This was somewhat surprising, given that the external cavity systems using KTP and KTA reported in [19,21] had generated THz output as high as 6.13 and 6.43 THz, respectively. To explore this further, we now consider the various factors relating to the generation of THz frequencies above 4.15 THz in RTP, including the efficiency with which it can be extracted, and the efficiency with which it propagates to, and is detected by the Golay cell.
4.1 THz generation efficiency
An indicator of how efficiently the THz field is being generated is to monitor the fundamental and Stokes output power across the tuning range. For each fundamental photon experiencing SPS, we assume that one Stokes and one THz photon must be generated. Therefore, the Stokes laser output power is, in theory, linearly proportional to the number of generated THz photons, in the absence of other processes. The Stokes laser output power and fundamental field depletion versus Stokes wavelength were measured, and are shown in Fig. 8, with the horizontal dashed line representing the Stokes output power detected at 1078.5 nm, the Stokes wavelength corresponding to 3.8 THz, which is the frequency at which the highest power was detected.
It can be observed in Fig. 8 that the detected Stokes output power is greater than that measured at 1078.5 nm for almost all Stokes wavelengths. This result suggests that the number of THz photons generated should be similar to or exceed that at 3.8 THz across the entire tuning range. Fundamental field depletion is high over the same range, with values between 50% and 55% until 1086.7nm, which corresponds to output at 5.7 THz. As described above, we expect that the THz power should scale similarly to the Stokes power, yet this is clearly not what we detect from our system. This suggests that there are other factors affecting our measurement of the THz field above 4.15 THz.
4.2 THz detection efficiency
The THz field is generated inside the nonlinear crystal, and travels through multiple materials and interfaces in order to be coupled out and detected. The efficiency with which THz photons are extracted from the laser and propagated to the Golay cell for power measurement is affected by THz absorption and Fresnel reflection losses associated with the various media in their path. Hence it is important to analyze the transmission for each step individually, starting with THz extraction from the laser, which involves propagation through the RTP and the silicon prisms), propagation through free space, and finally propagation through the detection optics in front of the Golay cell.
The first step in extraction is the propagation of the THz photons generated within the RTP to the crystal edge (to which the silicon prims are previously bonded). This THz field generated inside the crystal is outside the transparency range of RTP (0.35 - 4.5 µm), and therefore strong absorption of the THz field during propagation inside the crystal is expected. To quantify this, a direct measurement of the absorption coefficient in RTP at different THz frequencies was performed using the same method described in . In this technique, THz output power is measured versus THz field propagation distance inside the nonlinear crystal, and absorption coefficient is obtained according to Beer’s Law. The RTP crystal was mounted on a translation stage having micron-scale resolution, and the distance travelled by the THz beam inside the crystal was varied by translating the RTP across the fundamental beam propagation direction. At each position, the detected THz signal was recorded, and found to follow an exponential dependence, as shown in Fig. 9.
The absorption coefficients obtained from the exponential decay curves were 86 cm −1 at 3.80 THz, 87 cm−1 at 3.98 THz and 85 cm−1 at 4.10 THz, with estimated uncertainties of 10 cm−1. No values could be found in the literature for comparison. As expected, the material exhibits high absorption at THz frequencies, and minimizing the THz field propagation distance inside the RTP is therefore crucial. To this end, we always ensure that the fundamental beam resonates as close to the RTP crystal edge as possible. However, in order to avoid clipping of the Stokes beam, the minimum distance between the fundamental laser beam and RTP edges is around 500 µm. Taking this to be the average THz propagation distance inside the crystal, we find that more than 98% of the generated THz photons are absorbed inside the RTP for these three THz frequencies. It is difficult to predict the absorption losses for higher frequencies due to the shortage of reported material data. We expect, however, the absorption coefficient to generally increase as the THz frequency increases towards the strong 269 cm−1 resonance, with additional absorption in the vicinity of the weaker 142.0, 159, 163 and 211 cm−1 modes.
After propagation inside RTP, the transmitted THz photons are extracted with the silicon prisms, a process consisting of refraction in the crystal-prism interfaces, propagation inside the prisms, and a second refraction at the prism-air interfaces. Absorption in the high resistivity silicon prisms is known to increase considerably at higher THz frequencies , and therefore losses in the range of 10% at 4 THz are expected, increasing to 25% at 7 THz, given the 2 mm average propagation distance for the prisms used. At the crystal-prism and prism-air interfaces, the combined Fresnel reflection loss is estimated to be approximately 30% from 3 to 7 THz, given refractive indices for Si of 3.41, and assuming that for RTP it would be similar to that reported for KTP .
After being refracted out of the prism, the THz beam propagates in free space, and could experience absorption in air, particularly in water vapor. However, these discrete lines would generate dips and fine structure in THz spectra similar to what is observed in Fig. 4, and would not suppress a wide THz frequency band.
Finally, the THz beam impinges on the detection system, which is composed of a long pass filter, a TPX lens, and the Golay cell TPX entrance window. Each of these materials has a transmission loss that increases substantially with THz frequency . Table 1 lists the estimated transmission through each component, and it is seen that the total transmission (31%) for 4 THz radiation is around 10 times higher than the total transmission (3%) at 7 THz. As a consequence, the minimum detectable power depends strongly on the THz frequency. Given that the minimum detectable power at 3.8 THz was ~2 µW, we estimate the minimum detectable power would increase with frequency, up to 20 µW at 7 THz.
We believe that the absence of a detectable THz signal above 4.15 THz in our system is due to a combination of the expected increase in the absorption coefficient in RTP at higher frequencies (due to the tuning towards the strong absorbing mode at 269 cm−1), and the increasing losses in the outcoupling prisms and detection system with frequency above 4 THz. These effects attenuate the THz signal to a level which is below the minimum detectable signal level. A reduction in total transmission losses in the system is therefore necessary, and could be achieved either via optimizing the THz extraction from the RTP crystal, or reducing the losses in the detection system.
It is interesting to compare our findings to those in [19–21] where THz output above 6 THz was observed. First we note the higher pulse energies employed in [19–21], however the other significant difference was the surface-emitted configuration employed. In that configuration, the nonlinear crystal is cut at an angle to avoid total internal reflection of the THz beam at the crystal-air interface, enabling the THz field to be refracted out of the crystal without the silicon prisms. In the surface-emitted configuration the THz field is generated at the crystal-air interface, reducing absorption losses due to THz propagation inside the nonlinear material, and eliminating losses associated with Si prisms. An additional benefit is that the Golay cell can be positioned close to the nonlinear crystal, eliminating the need for a focusing lens, with its associated loss. In future work, we will explore the use of such a surface-emitting configuration in our intracavity system.
A frequency-tunable and high power THz source based on intracavity stimulated polariton scattering in RTP is demonstrated for the first time to the best of our knowledge. Frequency tunable THz output was obtained from 3.10 to 4.15 THz, with a gap from 3.17 to 3.49 THz, corresponding to the 104 cm−1 absorption mode in RTP. High output power levels were obtained, with a maximum average output power of 16.2 µW being detected at 3.80 THz, higher than that previously reported for an intracavity THz laser based on SPS. Our ongoing research aims overcome problems in extracting and detecting THz output above 4 THz, and thereby extend the tuning range of the system to around 6 THz. This work represents a compelling advance in the development of reliable, cost-effective and compact THz sources for real-world applications.
This work was supported under Australian Research Council’s Linkage Projects funding scheme (project number LP140100724) and was performed in part at the OptoFab node of the Australian National Fabrication Facility (ANFF), using NCRIS and NSW state government funding. Assoc. Prof. Pask is the grateful recipient of an Australian Research Council Future Fellowship (project number FT120100294), and Tiago Ortega gratefully acknowledges receipt of a scholarship from the Commonwealth of Australia under the International Postgraduate Research Scheme (IPRS). The authors would like to thank M Squared Lasers Inc. for their support.
References and links
1. P. H. Siegel, “Terahertz technology in biology and medicine,” IEEE Trans. Microw. Theory Tech. 52(10), 2438–2447 (2004). [CrossRef]
2. G. Davies, A. D. Burnett, W. Fan, E. H. Linfield, and J. E. Cunningham, “Terahertz spectroscopy of explosive and drugs,” Mater. Today 11(3), 18–26 (2008). [CrossRef]
3. R. M. Woodward, B. E. Cole, V. P. Wallace, R. J. Pye, D. D. Arnone, E. H. Linfield, and M. Pepper, “Terahertz pulse imaging in reflection geometry of human skin cancer and skin tissue,” Phys. Med. Biol. 47(21), 3853–3863 (2002). [CrossRef] [PubMed]
6. K. B. Cooper, R. J. Dengler, N. Llombart, T. Bryllert, G. Chattopadhyay, E. Schlecht, J. Gill, C. Lee, A. Skalare, I. Mehdi, and P. H. Siegel, “Penetrating 3-D imaging at 4- and 25-m range using submillimeter-wave radiation,” IEEE Trans. Microw. Theory Tech. 56(12), 2771–2778 (2008). [CrossRef]
8. E. Brundermann, H. Hubers, and M. F. Kimmitt, Terahertz Techniques (Springer, 2012).
9. K. Kawase, J. Shikata, and H. Ito, “Terahertz wave parametric source,” J. Phys. D Appl. Phys. 35(3), R1–R14 (2002). [CrossRef]
10. S. S. Sussman, Tunable Light Scattering from Transverse Optical Modes in Lithium Niobate (Stanford Univ., 1970).
11. K. Kawase, M. Sato, T. Taniuchi, and H. Ito, “Coherent tunable THz-wave generation from LiNbO3 with monolithic grating coupler,” Appl. Phys. Lett. 68(18), 2483–2485 (1996). [CrossRef]
12. K. Kawase, H. Minamide, K. Imai, J. Shikata, and H. Ito, “Injection-seeded terahertz parametric generator with wide tunability,” Appl. Phys. Lett. 80(2), 195–197 (2002). [CrossRef]
14. D. Walsh, S. J. M. Stothard, T. J. Edwards, P. G. Browne, C. F. Rae, and M. H. Dunn, “Injection-seeded intracavity terahertz optical parametric oscillator,” J. Opt. Soc. Am. B 26(6), 1196–1202 (2009). [CrossRef]
16. W. Wang, X. Zhang, Q. Wang, Z. Cong, X. Chen, Z. Liu, Z. Qin, P. Li, G. Tang, N. Li, C. Wang, Y. Li, and W. Cheng, “Multiple-beam output of a surface-emitted terahertz-wave parametric oscillator by using a slab MgO:LiNbO₃ crystal,” Opt. Lett. 39(4), 754–757 (2014). [CrossRef] [PubMed]
17. D. J. M. Stothard, T. J. Edwards, D. Walsh, C. L. Thomson, C. F. Rae, M. H. Dunn, and P. G. Browne, “Line-narrowed, compact, and coherent source of widely tunable terahertz radiation,” Appl. Phys. Lett. 92(14), 141105 (2008). [CrossRef]
18. J. Shikata, K. Kawase, K. Karino, T. Taniuchi, and H. Ito, “Tunable terahertz-wave parametric oscillators using LiNbO3 and MGO:LiNbO3 crystals,” IEEE Trans. Microw. Theory Tech. 48(4), 653–661 (2000). [CrossRef]
19. W. Wang, Z. Cong, X. Chen, X. Zhang, Z. Qin, G. Tang, N. Li, C. Wang, and Q. Lu, “Terahertz parametric oscillator based on KTiOPO₄ crystal,” Opt. Lett. 39(13), 3706–3709 (2014). [CrossRef] [PubMed]
20. C. Yan, Y. Wang, D. Xu, W. Xu, P. Liu, D. Yan, P. Duan, K. Zhong, W. Shi, and J. Yao, “Green laser induced terahertz tuning range expanding in KTiOPO4 terahertz parametric oscillator,” Appl. Phys. Lett. 108(1), 011007 (2016). [CrossRef]
21. W. Wang, Z. Cong, Z. Liu, X. Zhang, Z. Qin, G. Tang, N. Li, Y. Zhang, and Q. Lu, “THz-wave generation via stimulated polariton scattering in KTiOAsO4 crystal,” Opt. Express 22(14), 17092–17098 (2014). [CrossRef] [PubMed]
22. Y. Y. Wang, Z. X. Li, J. Q. Li, C. Yan, T. N. Chen, D. G. Xu, W. Shi, H. Feng, and J. Q. Yao, “Energy scaling of a tunable terahertz parametric oscillator with a surface emitted configuration,” Laser Phys. 24(12), 125402 (2014). [CrossRef]
23. A. Hildebrand, F. Wagner, J. Natoli, M. Commandre, H. Albrecht, and F. Theodore, “Laser damage investigation in RbTiOPO4 crystals: a study on the anisotropy of the laser induced damage threshold,” Proc. SPIE 6403, 64031W (2006). [CrossRef]
24. J. Y. Wang, L. Li, S. Liu, Y. Liu, and J. Wei, “The inelastic light scattering of RbTiOAsO4 single crystal,” Ferroelectrics 132(1), 197–202 (1992). [CrossRef]
25. P. Dekker, H. M. Pask, D. J. Spence, and J. A. Piper, “Continuous-wave, intracavity doubled, self-Raman laser operation in Nd:GdVO4 at 586.5 nm,” Opt. Express 15(11), 7038–7046 (2007). [CrossRef] [PubMed]
27. I. Hosako, N. Sekine, K. Fukunaga, Y. Kasai, P. Baron, T. Seta, J. Mendrok, S. Ochiai, and H. Yasuda, “At the dawn of a new era in terahertz technology,” Proc. IEEE 95(8), 1611–1623 (2007). [CrossRef]
28. A. Lee, Y. He, and H. Pask, “Frequency-tunable THz source based on Stimulated Polariton Scattering in Mg:LiNbO3,” IEEE J. Quantum Electron. 49(3), 357–364 (2013). [CrossRef]
29. V. Nikolaev, Deputy General Manager, Tydex LLC, 16 Domostroietelnaya str, St. Petersburg, Russia 194292 (personal communication, 2015).