Average-power scalability of multi-cycle terahertz sources based on periodically poled lithium niobate stacks Journal: opticajournal
Patrick J. Dalton, Robin Löscher, Tim Vogel, Robert B. Appleby, Graeme Burt, Steven P. Jamison, Morgan T. Hibberd, Darren M. Graham, Clara J. Saraceno · 2025
Read Average-power scalability of multi-cycle terahertz sources based on periodically poled lithium niobate stacks Journal: opticajournal by Patrick J. Dalton, Robin Löscher, Tim Vogel, Robert B. Appleby, Graeme Burt, Steven P. Jamison, Morgan T. Hibberd, Darren M. Graham, Clara J. Saraceno (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2509-13060.txt.
Introduction
Patrick J. Dalton and 8 co-authors show how to raise the average power of multi-cycle terahertz sources in a 2025 paper. Their sources rest on stacks of lithium niobate wafers. The wafers are periodically poled, a treatment that lets the crystal produce narrowband terahertz radiation.
The driving light came from femtosecond ytterbium lasers, firing at high repetition rates with high energy. On a stack of ten wafers, at a repetition rate of ten kilohertz, the team measured the average terahertz power delivered. Bigger stacks told a cautionary story. In stacks of forty-seven wafers, the beam bent sharply as it passed through, an effect like a lens. The authors measured that bending carefully and named it the main limit on the present setup. They offer ways to soften the problem in future designs.
Catalog record
| CREATOR | Patrick J. Dalton, Robin Löscher, Tim Vogel, Robert B. Appleby, Graeme Burt, Steven P. Jamison, Morgan T. Hibberd, Darren M. Graham, Clara J. Saraceno |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 18 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2509-13060
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Average-power scalability of multi-cycle terahertz sources based on periodically poled lithium niobate stacks Journal: opticajournal Patrick J. Dalton, Robin Löscher, Tim Vogel, Robert B. Appleby, Graeme Burt, Steven P. Jamison, Morgan T. Hibberd, Darren M. Graham, Clara J. Saraceno Abstract We demonstrate that narrowband multi-cycle terahertz (MC-THz) sources based on periodically-poled lithium niobate (PPLN) wafer stacks can be driven by high repetition-rate, high energy femtosecond ytterbium-doped lasers. Operating at 10-kHz repetition rate with up to 104 W of pump power on a 10-wafer stack, we measure 26.4 mW of THz average power for a narrowband multi-cycle source. We identify and quantify strong lensing effects causing dramatic beam focusing in 47 wafer stacks which act as a primary limitation in the current configuration, and present mitigation strategies for future scaling. This first study of high average power narrowband multi-cycle THz sources offers a path forward to Watt-level high repetition rate sources using thin lithium niobate plates. Narrowband multi-cycle THz (MC-THz) sources are of interest for a large number of applications including the selective excitation of resonant modes in materials [ 1 ] , harmonics generation [ 2 ] , and compact particle accelerators [ 3 ] . Growing interest has driven many recent advances [ 4 ] , with efforts focusing mainly on pulse energy scaling [ 5 ] . Another important parameter with relevance to the applications of these sources is their repetition rate. Higher repetition rates enable faster THz transient acquisition, rapid statistics, and improved experimental efficiency, particularly in electron acceleration [ 6 ] . However, little effort has been made to make use of high average power ytterbium (Yb) drivers to increase the repetition rate of these narrowband sources beyond 1 kHz. Most efforts so far to increase average power of THz sources have concentrated on single-cycle broadband THz sources [ 7 ] . The most commonly used approach for generating MC-THz pulses is nonlinear optical down-conversion of femtosecond lasers in large aperture, periodically-poled lithium niobate (PPLN). Two types of approaches have been studied so far to reach high energies: traditional voltage-poled bulk material [ 8 ] , and stacks of large aperture sub-millimeter thick wafers in alternating orientations [ 9 ] . Stacked lithium niobate (LN) plates are particularly interesting due to their tunability by the number of wafers stacked, i.e. the pump beam interaction length, to control the number of cycles and tailor the generation process for the desired applications [ 10 ] . Furthermore, much larger aperture wafers are available than bulk crystals, allowing straightforward energy scaling with large apertures. A comparison of both techniques has been recently shown in [ 5 ] using PPLN and KTP, highlighting the challenges related to very high energy excitation in LN, such as multi-photon absorption, infrared induced absorption and others. At high average powers, these challenges are known to become more severe due to cumulative effects and additional thermal management difficulties. For example, the thermal load at elevated repetition rates will result in changes of refractive index and absorption of the material in both the optical and THz beams [ 11 , 12 ] , which leads to effects such as thermal lensing or depolarization. Although the thermo-optic and thermal conductivity coefficients in LN are well known, the impact of these effects for high average power THz sources have only recently started to be studied in the context of the titled-pulse front technique for broadband single cycle pulses [ 7 ] , and there are no reports of increasing the repetition rate of PPLN-based MC-THz sources beyond 1 kHz to explore the influence of thermal effects. In this letter, we generate narrowband MC-THz radiation in a PPLN wafer stack (PPLN-WS) driven by a Yb-based thin-disk amplifier laser system operating at 1030-nm wavelength, 10-kHz repetition rate, and 700-fs pulse width. By pumping the PPLN-WS with 104 W of average power, we measure 26.4 mW of THz average power. We identify a combination of photorefractive-, thermal-, and Kerr-lensing in the PPLN as the main limitation for our approach, warranting further studies on the interplay of pump peak and average power for future Watt-class MC-THz sources. Our work presents first steps towards improved measurement statistics and rapid experiments, for example in laser-driven particle acceleration at higher repetition rates. Figure 1: Schematic diagram of the THz generation and detection setup. HWP = half-wave plate, BS = beam splitter, OAP = off-axis parabolic mirror, QWP = quarter-wave plate, WP = Wollaston prism, PD = balanced photodetector, BP Filter = bandpass filter, PM = power meter, BD = beam dump. Figure 2: Unconverted pump beam profiles measured after transmission through a PPLN stack (47 wafers) for average pump powers between a) 22 W and f) 66 W. Corresponding x - and y -line profiles in subfigures g) to l). Experimental setup. In this experiment we used 5-mol.% magnesium oxide-doped congruent x -cut LN wafers with a diameter of 50.8 mm and a nominal thickness of 135 µm with an anti-reflection coating for the pump wavelength. The wafer thickness corresponded to a poling period of 270 µm, and a THz output frequency around 0.4 THz in the forward propagation direction. A 500-µm thick z -cut quartz wafer was added to reduce Fresnel losses at the LN-vacuum interface for optimized out-coupling of the THz radiation. The driving laser is a thin-disk regenerative amplifier system (DIRA 500-10, TRUMPF Scientific Lasers) and the experimental setup is shown in Fig. 1 . The pump average power applied to the PPLN was controlled by a half-waveplate and a thin-film polarizer with calibrated average power in front of the PPLN-WS. The output was split into a pump and gate pulses for electro-optic sampling (EOS) by an output coupler with 0.2% transmission. The pump beam area was reduced to 6.04 \times 7.51 mm (1/ \mathrm{e}^{2} ) using a Galilean mirror-based telescope. Two 90 ∘ off-axis parabolic mirrors (OAP) with focal lengths of 101.6 mm, and central 3-mm diameter holes, were used to collect the generated THz radiation and focus it onto either a (110)-cut 1-mm thick gallium phosphide (GaP) crystal for EOS, or a thermopile sensor (3A-P-THz, Ophir) for THz average power measurements. The hole in the first parabolic mirror allowed the unconverted pump power to be directed into a beam dump, while the hole in the second parabolic mirror allowed for EOS with a collinear probe beam. A THz transmitting filter was placed in the collimated THz beam path between the two OAPs to block residual 1030-nm pump light and any 515-nm second-harmonic generated in the PPLN-WS. The filter was made from a combination of polytetrafluoroethylene (PTFE) and high-density polyethylene (HDPE) plates. Values for the THz average power within this letter are corrected for reflective losses at the filters and transmission losses at the holes of the OAPs. The EOS measurements employed a sinusoidally moving delay line to provide a 4̃0-ps delay range for the gate pulse and an open-source algorithm was used for data processing [ 13 ] . An additional motorized delay stage allowed the acquisition of THz waveforms over longer time delays by stitching individual 40-ps slices. The stitching algorithm utilized a correlation function for optimized delay shifting. The unconverted pump beam profile exiting the PPLN-WS was measured