Mid-infrared (IR) lasers are currently an area of rapid development, with several competing technologies. In traditional gas lasers, the effective interaction length is limited and the system as a whole is bulky and inflexible, limiting their applications. Standard gain fibers cannot be used in the mid-IR because the glass forming the fiber core is not transparent at these longer wavelengths. In this Letter, we report the demonstration of a mid-IR fiber gas laser using feedback in an optical cavity. The laser uses acetylene gas in a high-performance silica hollow-core fiber as the gain medium, and lases either continuous wave or synchronously pumped when pumped by telecom-wavelength diode lasers. We have demonstrated lasing on a number of transitions in the spectral band of 3.1–3.2 μm. The system could be extended to other selected molecular species to generate output in the spectral band up to 5 μm, and it has excellent potential for power scaling.
Published by The Optical Society under the terms of the Creative Commons Attribution 4.0 License. Further distribution of this work must maintain attribution to the author(s) and the published article's title, journal citation, and DOI.
Mid-infrared (IR) represents a rapidly evolving frontier for laser development. Quantum cascade lasers have steadily extended through this band, but continuous-wave (CW) room-temperature emissions between 3 and 3.5 μm remain elusive . The spectacular success of silica fiber lasers  cannot be extended much beyond the 2.5 μm wavelength, because silica becomes strongly absorbing at these longer wavelengths. Alternative glass systems such as chalcogenides have made great progress , but have drawbacks such as fiber strength and stability, nonlinearity, and optical damage. Systems based on nonlinear conversion  are versatile but typically bulky and inefficient. We now demonstrate that a gas-filled silica-based hollow-core fiber laser enables both CW and synchronously pumped mid-IR (3.1 μm) lasing with a sufficiently low threshold to be directly optically pumped by an unamplified () 1.5 μm diode laser, none of which has been demonstrated before in this spectral band. This system is convenient, robust, exhibits a low pump threshold, and has tremendous potential for scalability in footprint, power, and wavelength in the mid-IR. Our interest is in creating a versatile laser technology for the 3–5 μm spectral band that is compact, rugged, and scalable to high optical powers. Hollow-core silica fibers combine the strength and durability of silica with record low attenuation in the 3–4 μm spectral band, where their attenuation can be 10,000 times less than that of bulk silica [5–7] and the equal of any conventional optical fibers [8,9]. They also have a very high damage threshold [10,11], very low group-velocity dispersion , and an optical nonlinearity that is many orders of magnitude less than that of competing fibers . They thus have the potential to substantially outperform other fibers in the mid-IR. At wavelengths beyond 3 μm, softer glasses like chalcogenides are considered to have outstanding potential as hosts for rare earth elements. However, these glasses are hard to purify, are not always stable in ambient conditions, and difficult to draw into fibers. The silica-based hollow fibers used in our experiments offer a complementary approach to fiber lasers in the mid-IR, and are shown in Fig. 1. Further information is in Table S1 of Supplement 1.
Our work builds on recent research showing how molecular species such as acetylene and HCN can provide gain at mid-IR wavelengths at low vapor pressure in a cavity when pumped with nanosecond pulses from an optical parametric converter , and how this system transfers into a hollow-core fiber in a single-pass amplified-spontaneous-emission (ASE) configuration . A system using acetylene as the gain medium can be pumped using a stabilized, modulated, fiber-amplified diode laser around 1530 nm . In those experiments, although amplification of spontaneous emissions was enhanced by the waveguiding provided by the fiber, there was no cavity feedback and only pulsed single-pass operation was reported. Pump peak powers of tens of watts were required for appreciable conversion [15,16], while CW operation was not achieved. CW operation of an acetylene laser is made harder by the lack of a radiative transition from the lower lasing levels. The requirement for a nonradiative (collisional) de-excitation pathway affects the laser dynamics and limits the efficiency . Recently, the CW operation of a different molecular system at 1310 nm was reported . A linear cavity incorporating a short (20 cm) straight hollow fiber placed within a gas cell containing molecular iodine was pumped with a tunable, single-frequency, frequency-doubled pump source. Very high fiber attenuation at the pump wavelength () was overcome by using only a short fiber length and high gain from a molecular transition with rapid de-excitation of the lower laser level.
Our system (Fig. 2) relies on the use of a hollow-core fiber in two ways. The first is to simultaneously confine the light (at both pump and laser wavelengths) and the molecular gain medium. The attenuation of the (empty) gain fiber at the pump wavelength of 1.53 μm is 0.11 dB/m and the attenuation around 3.1–3.2 μm, the laser wavelength band, is 0.10 dB/m. Our low-loss fiber enables the use of low gas pressure to provide a very high gain per pass. By using different gain fiber lengths, we found that 10 m gives the best laser performance. The gain fiber is enclosed at each end in a small gas cell, providing optical access through a window. The cells and fiber were evacuated and then filled with a low pressure (typically, 0.3 mBar) of acetylene.
Our feedback fiber [Figs. 1(b) and 3(a)] has one of the lowest attenuations reported for any optical fiber in this spectral band, enabling long optical cavities and, hence, low-repetition-rate pulsed lasing. Acetylene is pumped using a CW distributed feedback diode laser tuned to one of the absorption lines around 1530 nm as shown in Fig. 3(b), which can be modulated and/or amplified using telecom technology when required. For pulsed operation, the pump repetition rate is carefully adjusted (using Stanford Digital Delay Generator DG645 to control the pump modulator) so that the circulating laser pulse is coincident in time with the 1.53 μm pump pulse. A dichroic mirror (high transmission at 1.53 μm, high reflection at 3.16 μm) is used to recombine the feedback with the pump before coupling into the gain fiber. Our use of a feedback fiber as part of a ring cavity configuration facilitates cavity alignment. A short feedback fiber (e.g., 3 m) provides lowest loss for CW operation and a longer length (100 m) allows pulsed operation at a relatively low repetition rate. About 7% or 70% of the total power is coupled out of the cavity per round trip using an uncoated window or a coated custom-made optical component, respectively, as the output coupler.
In operation, the pump was tuned to one of the acetylene P-branch absorption lines around 1530 nm. We used erbium-doped fiber amplifiers to increase the pump power to around 75 mW, and coupled around 80% of this into the gain fiber through the dichroic mirror. With a feedback fiber length of 3 m, we then detected lasing at corresponding transitions around 3.1 μm, as illustrated in Fig. 4(a).
For CW pumping [Fig. 4(b)], we typically observed two lines arising from the transitions from the common upper level populated by the pump excitation. For each pump transition, the laser transitions share a common upper level  and so are in competition, and one might expect to observe just a single line above the threshold. However, the gain for the above-threshold line is reduced as the population builds up in its lower lasing level, until it is exceeded by the gain provided to the competing line (with a lower lasing level that is still empty), allowing this second line to lase as well. Thus, broadly up to half of the pump photons can result in a laser photon for the highest-gain line, and up to half of the remainder can result in a photon on the second line. Highest output powers are obtained using 70% output coupling as in Fig. 5(a). Under CW pumping, our cavity oscillated bi-directionally, with slope efficiency of 6.7% (single output measured). No saturation of the output power was observed as we increased the pump power to maximum (75 mW). To reduce the threshold, we have used a lower output coupling of 7%. A reduced pump threshold of 16 mW then enables our cavity to lase when pumped directly by our unamplified pump laser, although with reduced efficiency. This configuration is ideal for use as a seed in a possible oscillator/amplifier configuration, with overall amplifier efficiency expected to be . Figure 5(b) shows the high beam quality of the laser output and the excellent stability of the laser over long time scales.
Replacing the feedback fiber with a longer fiber of 100 m length and modulating the pump into a train of 80 ns pulses causes a strong dependence of the 3.1 μm laser output on the repetition rate, as shown in Fig. 6. Maximum output power is recorded when the repetition rate is at the frequency of the ring cavity (2.6 MHz—synchronous pumping). When the pump repetition rate is changed, the lasing continues, although with reduced efficiency and stability. When the feedback is blocked, no output can be detected. The laser power recovers around harmonic and subharmonic repetition rates. As the pump repetition rate is reduced (but without adjusting amplifier pumps), the pump pulse energy increases and less feedback is required to generate a detectable output, resulting in less sensitivity to the pump repetition rate as we approach the single-pass ASE regime. However, we still are unable to detect an output under any circumstances when the feedback is blocked because even at these lower repetition rates (0.1 MHz), the pump peak powers (5.8 W) and pulse energies (0.46 μJ) are much lower than in previous single pass ASE experiments . Increasing the repetition rate above the cavity frequency, there is a rapid decrease of the output, indicating that the decay of the excited state is rapid even at low pulse energies. Spectrally, the output when synchronously pumped is dominated by a single strong line, with only the very weak detection of a second line. Using an average pump power of tens of milliwatts (peak power at 2.6 MHz) we easily reach the threshold for laser oscillation under pulsed pumping and our cavity oscillates with a slope efficiency of 8.8% as shown in Fig. 7. The threshold under synchronous pulsed pumping is not affected by reducing the output coupling, indicating that the gain is saturated. Just as for CW operation, the output was stable over many hours (see Fig. S1 in Supplement 1).
Emissions from the system when synchronously pumped take the form of a pulse train emitted at the round-trip repetition rate of the laser cavity, and correspond to copies of a single pulse circulating inside the cavity (see Fig. S2 in Supplement 1). Radio-frequency (RF) spectra of our laser are shown in Fig. 8(a) along with the optical spectrum of the laser [Fig. 8(b)] and the temporal profiles of the pump and laser pulses [Fig. 8(c)], for selected frequencies. When pumped synchronously (2.6 MHz), the RF spectra reproduce the noise characteristics of the modulated pump source: at other repetition rates, they reveal the complex dynamics of asynchronous pumping. Further data are in Table S2 of Supplement 1.
In conclusion, we report a diode-laser-pumped ring-cavity all-silica fiber gas laser with a low threshold running stably around 3 μm wavelength either CW or pulsed with a repetition rate of 2.6 MHz. The laser is enabled by the remarkable properties of the mid-IR hollow-core fiber formed from silica, and is pumped around 1530 nm wavelength. The system is scalable through the use of existing 1.5 μm laser technology using an oscillator/amplifier configuration, and could be extended to different wavelengths using other molecular gain systems up to 5 μm wavelength.
All data underlying the results presented in this Letter can be found at .
Engineering and Physical Sciences Research Council (EPSRC) (EP/I011315/1); Majlis Amanah Rakyat (MARA).
See Supplement 1 for supporting content.
1. Y. Yao, A. J. Hoffman, and C. F. Gmachl, Nat. Photonics 6, 432 (2012). [CrossRef]
2. D. J. Richardson, J. Nilsson, and W. A. Clarkson, J. Opt. Soc. Am. B 27, B63 (2010). [CrossRef]
3. S. D. Jackson, Nat. Photonics 6, 423 (2012). [CrossRef]
4. R. A. Kaindl, M. Wurm, K. Reimann, P. Hamm, A. M. Weiner, and M. Woerner, J. Opt. Soc. Am. B 17, 2086 (2000). [CrossRef]
5. F. Yu, W. J. Wadsworth, and J. C. Knight, Opt. Express 20, 11153 (2012). [CrossRef]
6. F. Yu and J. C. Knight, Opt. Express 21, 21466 (2013). [CrossRef]
7. N. V. Wheeler, A. M. Heidt, N. K. Baddela, E. N. Fokoua, J. R. Hayes, S. R. Sandoghchi, F. Poletti, M. N. Petrovich, and D. J. Richardson, Opt. Lett. 39, 295 (2014). [CrossRef]
8. J. Bei, T. M. Monro, A. Hemming, and H. Ebendorff-Heidepriem, Opt. Mater. Express 3, 1285 (2013). [CrossRef]
9. W. H. Kim, V. Q. Nguyen, L. B. Shaw, L. E. Busse, C. Florea, D. J. Gibson, R. R. Gattass, S. S. Bayya, F. H. Kung, G. D. Chin, R. E. Miklos, I. D. Aggarwal, and J. S. Sanghera, J. Non. Cryst. Solids 431, 8 (2016). [CrossRef]
10. P. Jaworski, F. Yu, R. R. J. Maier, W. J. Wadsworth, J. C. Knight, J. D. Shephard, and D. P. Hand, Opt. Express 21, 22742 (2013). [CrossRef]
11. P. Jaworski, F. Yu, R. M. Carter, J. C. Knight, J. D. Shephard, and D. P. Hand, Opt. Express 23, 8498 (2015). [CrossRef]
12. M. G. Welch, C. E. de Nobriga, R. A. Correa, W. J. Wadsworth, and J. C. Knight, Opt. Express 17, 9006 (2009). [CrossRef]
13. D. G. Ouzounov, F. R. Ahmad, D. Müller, N. Venkataraman, M. T. Gallagher, M. G. Thomas, J. Silcox, K. W. Koch, and A. L. Gaeta, Science 301, 1702 (2003). [CrossRef]
14. A. V. V. Nampoothiri, A. Ratanavis, N. Campbell, and W. Rudolph, Opt. Express 18, 1946 (2010). [CrossRef]
15. A. M. Jones, A. V. V. Nampoothiri, A. Ratanavis, T. Fiedler, N. V. Wheeler, F. Couny, R. Kadel, F. Benabid, B. R. Washburn, K. L. Corwin, and W. Rudolph, Opt. Express 19, 2309 (2011). [CrossRef]
16. Z. Wang, W. Belardi, F. Yu, W. J. Wadsworth, and J. C. Knight, Opt. Express 22, 21872 (2014). [CrossRef]
17. J. Han, K. Freel, and M. C. Heaven, J. Chem. Phys. 135, 244304 (2011). [CrossRef]
18. A. V. V. Nampoothiri, B. Debord, M. Alharbi, F. Gérôme, F. Benabid, and W. Rudolph, Opt. Lett. 40, 605 (2015). [CrossRef]
19. A. V. V. Nampoothiri, A. M. Jones, C. Fourcade-Dutin, C. Mao, N. Dadashzadeh, B. Baumgart, Y. Y. Wang, M. Alharbi, T. Bradley, N. Campbell, F. Benabid, B. R. Washburn, K. L. Corwin, and W. Rudolph, Opt. Mater. Express 2, 948 (2012). [CrossRef]