Towards high power mid-infrared emission from

NATURE PHOTONICS | VOL 6 | JULY 2012 | www.nature.com/naturephotonics 423

Fibre lasers are efficient, powerful and versatile waveguide reso-nant devices that comprise glass optical fibre waveguides for optical gain and Fabry–Pérot resonators for optical feedback.

Rapid developments in fabrication capabilities now allow fibres composed of ultralow-loss silicate glass with low scattering, impurity losses and material imperfections, thus providing enormous flex-ibility in the characteristics and quantity of light that can be gener-ated from fibre lasers. Passively air-cooling an optical fibre is simple owing to its large surface-area-to-volume ratio. Optical excitation using multimode semiconductor diode lasers is straightforward and efficient with cladding pumping1 — particularly when the axial symmetry of the fibre is broken2 — and efficiently excites the single-mode core of the fibre to create near-diffraction-limited output and a brightness enhancement of at least three orders of magnitude3. Although activating silicate glass optical fibres using rare-earth cat-ion doping introduces additional Rayleigh scattering, this drawback is small compared with the gain enhancement. Furthermore, dilute (<0.1 mol%) rare-earth cation concentrations have a background loss4 of less than 2 dB km–1 — a value that is commensurate with conventional fibre. The lack of long-range order in a silicate glass means that the bandwidth for pump absorption and gain can be up to 50 THz, which relaxes the wavelength tolerance for excitation and creates opportunities for broad tuning and ultrashort pulse genera-tion. The small dimensions of the modes propagating in the active core and pump core of the fibre (Fig. 1) provide low-threshold and high-gain characteristics with near-maximum efficiency.

The creation and development of discrete and tunable sources of mid-infrared laser radiation has a rich history. Quantum cascade lasers5 are excellent mid-infrared light generators that currently form the basis of mid-infrared photonics. Quantum cascade lasers oper-ating in continuous-wave (CW) mode produce a large amount of heat, which must be dissipated from the active region for operation at room temperature. Calculations6 suggest that most (>70%) of the injected electrical power is converted into heat. Given that the area of the active region is approximately 100 μm2, it will be a challenge to create high-power single-mode infrared light conveniently from these sources. Vibronic solid-state lasers — that is, lasers displaying broad gain bandwidths arising from strong interactions with phon-ons, based on Cr2+ or Fe2+ cations doped into ii–vi compounds7 — have shown great progress for emission in the range of 2–5 μm. The efficiencies of these lasers are sufficient for commercial exploitation. Chalcogenide ceramic lasers8 based on Cr2+:ZnSe, which can produce output powers of 15 W in a single, CW longitudinal mode, are com-mercially available9, but thermal lensing is a challenge and quench-ing from multiphonon emission may limit power scaling to around

Towards high-power mid-infrared emission from a fibre laserStuart D. JacksonThe diverse output pulsewidths, linewidths and polarization states of fibre lasers provide both high efficiency and high output power in a small, low-maintenance format that is ideal for applications throughout research, defence and industry. New direc-tions are constantly being pursued to exploit the capabilities of this technology. One such direction involves extending the emission wavelength further into the infrared, which will benefit numerous existing and future applications. Many exciting advances have been demonstrated and many challenges remain, the most significant of which are summarized in this Review.

20 W (ref. 10). Although direct bandgap iii–v semiconductor diode lasers can operate in CW mode at room temperature in the range of 1.9–2.7 μm (ref. 11), Auger recombination, which limits the carrier lifetime, significantly reduces the efficiency at wavelengths longer than 1.8 μm. Commercial optical parametric oscillators that offer wide tunability12 and multiwatt output power in a robust, compact system are now available. However, parametric generation requires a narrow linewidth and linear polarized excitation — both of which are significant constraints on the pump13. Fibre lasers fit securely in this continuum of choice and offer some important advantages over optical parametric oscillators.

The highest values of output power and efficiency have been achieved at around 1 μm using Yb3+ cations doped into the core of silicate glass fibre14. Silicate glasses are a broad group of glasses, but the most relevant to fibre lasers involve aluminosilicates and germanosilicates. Commercial systems9 capable of supplying out-put powers of up to 50 kW are now used for a variety of applica-tions, including cutting and welding in the automotive industry. There are a number of reasons for this success. First, the quantum

Institute of Photonics and Optical Science and Centre for Ultra-broadband Devices and Optical Systems, School of Physics, University of Sydney, Camperdown, New South Wales 2006, Australia. e-mail: [email protected]

Active core

Pump core

Polymeric coating

Figure 1 | Cross-section of a typical rare-earth-doped double-clad fibre for use in a fibre laser. The active core is doped with rare-earth ions that provide the absorption and emission processes necessary for lasing. The pump core usually does not have a circular cross-section because this would limit absorption.

FOCUS | REVIEW ARTICLESPUBLISHED ONLINE: 28 JUNE 2012 | DOI: 10.1038/NPHOTON.2012.149

© 2012 Macmillan Publishers Limited. All rights reserved.

mailto:[email protected]

424 NATURE PHOTONICS | VOL 6 | JULY 2012 | www.nature.com/naturephotonics

efficiency of the laser transition is close to 100% while the quantum defect — that is, the difference between the pump and laser photon energies — is typically less than 10%. Second, the use of silicate glass optical fibre provides the system with a significant degree of robustness and power-handling capability. Third, the constraints on the Yb3+ concentration are relaxed because the ‘two level’ energy level structure mitigates energy transfer between the Yb3+ cations; energy transfer can be a source of excitation loss for other rare-earth cations. Yb3+-doped silicate glass fibre lasers emitting at around 1 μm are therefore incredibly practical devices, but extend-ing the emission wavelength from fibre lasers towards and into the mid-infrared will be necessary for a large number of existing and future applications.

The task of engineering a fibre laser becomes easier at longer wavelengths, owing to a number of optical scaling factors. In the weakly guiding approximation, the mode area of the lowest-order mode scales as λ2, where λ is the laser wavelength; for example, the LP01 mode at 2 μm has four times the area of the same mode at 1 μm. This has a profoundly beneficial effect on the power scaling potential of fibre lasers. Because losses due to nonlinearity scale with light intensity, Brillouin and Raman scattering are lower at longer wavelengths and so optical damage thresholds are increased. The size of the mode can also be increased by using smaller refrac-tive index contrasts15 or holey fibre16. Translating these methods to longer wavelengths provides further potential for power scaling.

Moving further into the infrared and away from the bandgap of a given glass means that more photons are required to bridge the bandgap, which raises the ablative threshold and thus lowers the loss from two-photon absorption. Stokes emission generated from Raman scattering is also weaker at longer wavelengths17. The forma-tion of colour centres resulting from localized charge regions is a problem for some Yb3+-doped silicate glasses because the resulting absorption features located in the visible region have long absorp-tion tails that overlap with the pump and emission wavelengths of Yb3+ (refs 18,19). The absorption strength diminishes and therefore potentially becomes less problematic at longer wavelengths; for example, with silicate glass fibre lasers that operate at 2 μm.

Unfortunately, extending the emission wavelength of fibre lasers beyond the natural loss-minimum of silicates at 1.5 μm remains a significant challenge. The physical properties of the glass com-prising some fibres, such as low values for background absorption, thermal conductivity and glass transition temperature, nega-tively impact the output performance beyond this loss minimum. Infrared transmission in the glass is controlled by the phonon den-sity of states, which means that the longest emitting wavelength from a fibre laser is always shorter than its maximum transmissible wavelength because the phonon density of states determines the radiative efficiencies of the fluorescence transitions of the rare-earth cations. Significant research has been directed towards the development of glasses that have the combined characteristics of low maximum phonon energy and fibre ‘drawability’ while main-taining amorphousness and low loss. Only a small number of glasses currently have these characteristics.

Table B1 | Characteristics of infrared fibre lasers with emission wavelengths ≥1.5 μm.

Dopant(s) Host glass Pump λ (μm) Laser λ (μm) Transition Output power (W) Slope efficiency (%) Reference

Er3+, Yb3+ Silicate 0.975 1.5 4I13/2 → 4I15/2 297 19 21

Tm3+, Ho3+ ZBLAN 0.792 1.94 3F4 → 3H6 20 49 33

Tm3+ Silicate 0.793 2.05 3F4 → 3H6 1,050 53 22

Tm3+, Ho3+ Silicate 0.793 2.1 5I7 → 5I8 83 42 34

Ho3+ Silicate 1.950 2.14 5I7 → 5I8 140 55 23

Tm3+ ZBLAN 1.064 2.31 3H4 → 3H5 0.15 8 35

Er3+ ZBLAN 0.975 2.8 4I11/2 → 4I13/2 24 13 24

Ho3+, Pr3+ ZBLAN 1.1 2.86 5I6 → 5I7 2.5 29 25

Dy3+ ZBLAN 1.1 2.9 6H13/2 → 6H15/2 0.275 4.5 36

Ho3+ ZBLAN 1.15 3.002 5I6 → 5I7 0.77 12.4 26

Ho3+ ZBLAN 0.532 3.22 5S2 → 5F5 0.011 2.8 27

Er3+ ZBLAN 0.653 3.45 4F9/2 → 4I9/2 0.008 3 28

Ho3+ ZBLAN 0.89 3.95 5I5 → 5I6 0.011 3.7 29

Out

put p

ower

(W)

105

104

103

102

101

100

10–1

10–2

10–3

Emission wavelength (µm)

Mid-infrared

Near-infrared

1.0 1.5 2.0 2.5 3.0 3.5 4.0

Yb3+ [20]

Tm3+ [22]

Er3+ [21]

Er3+ [24]

Er3+ [28]

Ho3+ [23]

Ho3+ [27] Ho3+ [29]

Ho3+ [25]

Ho3+ [26]

Figure 2 | Output power from published demonstrations of infrared fibre lasers as a function of the emitted wavelength. The reduction in output power with increasing emission wavelength is caused primarily by the growing quantum defect between the pump photon energy and laser photon energy. Cryogenic cooling of the fluoride fibre was needed to demonstrate emission at 3.9 μm (ref. 29).

REVIEW ARTICLES | FOCUS NATURE PHOTONICS DOI: 10.1038/NPHOTON.2012.149



Despite the variety of rare-earth cation transitions and host materials that have been tested for fibre laser emission, the maxi-mum CW output power produced from previously demonstrated fibre lasers has a clear exponential decrease when plotted as a function of emission wavelength (Fig. 2)20–29. The primary cause of this power drop is the increase in quantum defect at longer wave-lengths. Efficient high-power diode lasers — the traditional pump sources for silicate glass fibre lasers — typically emit light in the near-infrared region close to 1 μm. Although it would be ideal to use these efficient pump sources for all fibre lasers, the larger quantum defect at longer wavelengths creates heat that becomes an ever-increasing fraction of the absorbed pump energy. The small active-core volume of an optical fibre relative to the total fibre vol-ume of a double-clad fibre results in a small temperature gradi-ent30. The temperature within the fibre is governed strongly by the degree of cooling at the air–fibre boundary, which is improved by a large surface area (that is, a larger pump-core diameter). Forced air cooling provides a particular benefit over passive convective cooling30. The excellent cooling properties of optical fibres there-fore help to alleviate the problem of the growing quantum defect. Developing efficient longer-wavelength pump sources and select-ing more appropriate laser transitions can effectively reduce the quantum defect and improve output performance. Laser transi-tion cascading31, in which multiple transitions lase simultaneously after a single pump photon excitation, offers the opportunity for better photon-to-photon conversion efficiencies and is particularly suited to longer-wavelength transitions for reducing heat loads and improving power-scaling opportunities.

Laser transitions of the rare-earth cationsThe lasing wavelength of long-wavelength fibre lasers is determined by the fluorescence transitions of the rare-earth cation doped into the core of the fibre. Figure 3 shows the laser transitions responsible

for infrared emission from rare-earth-cation-doped fibre lasers. Although the nomenclature for describing the infrared region gen-erally depends on the research field, for clarity the divisions in this Review adhere to the International Standard32. Table B1 lists the max-imum output power produced from the infrared transitions of previ-ously demonstrated rare-earth-cation-doped fibre lasers21–29,33–36. To create silicate glass preforms that can then be drawn into fibres, rare-earth cations in solution substitute for H+ cations of silanol (SiOH) groups located at the surface of the silicate glass flocculent layer, which is deposited on the interior surface of the preform start tube during fabrication. With a few exceptions, rare-earth cations are con-sistently in the trivalent oxidation state, and the associated electronic transitions of the cations are the foundation of infrared fibre lasers. Although transition metal cations have been doped into glasses and fluoresce for emission at long wavelengths, transition metal ions have large emission probabilities compared with rare-earth cations and so their energy-level lifetimes are short. In non-crystalline host materi-als, transition metal cations can exist in a number of valence states that can introduce a number of effects to the lasing process, including parasitic pump absorption, reabsorption of the laser light, lifetime quenching of the lasing ion and trapping of the excitation energy. Phonon broadening acts similarly on all of the rare-earth cations and is therefore homogeneous, but the perturbations to the energy levels of the cations by the surrounding electric fields from nearby glass atoms vary from one cation site to another, which makes them inho-mogeneous and temperature-independent. In covalently bonded glasses such as silica, the rare-earth cations form inter-network regions37 (percolation channels) comprised of rare-earth cations that bond ionically to non-bridging oxygen atoms that interface with the main covalently bonded network regions. Because their large cati-onic field strength (defined as Z/r2, where Z is the atomic number and r is the radius of the ion) requires a high coordination num-ber, rare-earth cations tend to cluster in the inter-network regions to

1G4

3F2,3

4F7/2

5S2

5F5

5I4

5I5

5I6

5I7

5I86H15/2

6H13/2

6H11/2

6F11/26H9/2

6F9/26H7/2

2H11/24S3/2

4F9/2

4I9/2

4I11/2

4I13/2

4I15/2

3H4

3H5

3H6

Tm3+ Er3+ Ho3+ Dy3+

3F4

Pump ESAPump ESA

2.3 µm

3.45 µm

2.75 µm

1.55 µm2 µm 2.1 µm

2.9 µm

2.85 µm

3.95 µm

3.22 µm

Figure 3 | Laser transitions of rare-earth cations that produce emission wavelengths longer than 1.5 μm. Figure shows near-infrared (blue arrows), mid-infrared (red arrows) and both (green arrows) electronic transitions, together with the primary lasing wavelength from previously demonstrated fibre lasers for each transition. The fluorescence spectra of the 5I6 → 5I7 transition of Ho3+ and the 6H13/2 → 6H15/2 transition of Dy3+ extends across the boundary between the near- and mid-infrared regions. ESA, excited state absorption.

FOCUS | REVIEW ARTICLESNATURE PHOTONICS DOI: 10.1038/NPHOTON.2012.149



share non-bridging oxygen atoms. The nephelauxetic effect causes the absorption and emission peaks of rare-earth cation transitions to red-shift to longer wavelengths with increasing covalency of the bonding between the network formers38.

The fluorescence spectra of the near-infrared transitions from fibre lasers relevant to this Review are shown in Fig. 4. The quasi-three-level transitions of the large-Z rare-earth cations thulium iii, holmium iii and erbium iii, which fluoresce in the range of 1.5–2.2 μm, are responsible for the highest powers available from fibre lasers emitting in the λ > 1.5 μm region of the near-infrared. The output power performance of such devices relates to strong absorp-tion by the thulium, holmium and erbium cations, which overlaps with the emission from commercial off-the-shelf high power diode lasers and the excellent physical properties of the silicate glasses used to create the fibre. Fibre lasers employing the Tm3+ cation39–42 with emission at around 2 μm are the most powerful, efficient and developed of these fibre lasers. These devices are excited with estab-lished diode lasers emitting at 0.79 μm, and their output can be tuned from at least 1.86 μm to around 2.09 μm (ref. 43). With the implementation of Tm3+ concentrations exceeding 2.5 wt%, com-bined with cluster-reducing co-dopants that mitigate gain-lowering energy transfer upconversion processes, cross-relaxation between neighbouring Tm3+ cations can nearly double the slope efficiency44. Cross-relaxation (‘two-for-one’ excitation; Fig. 5) is resonant in sili-cate glass, thus providing one of the most efficient ways to generate 2 μm light from commercial diode laser pump sources. The 5I7 → 5I8 transition of Ho3+ has a peak emission wavelength of 2.1 μm, which overlaps with an important atmospheric transmission window. This transition is also acceptable for resonant pumping with Tm3+-doped silicate glass fibre lasers45, which introduces a small quantum defect of <7%. Ho3+ fibre lasers were initially46 co-doped with Tm3+ sen-sitizer cations to exploit the diode-pumpable absorption of Tm3+, although loss of excitation to energy-transfer upconversion and reversible energy-transfer between the first excited states of Tm3+ and Ho3+ limited the power, extractable energy and efficiency of this technique. Doping aluminosilicate glass with only Ho3+ cations sup-ports a large short-pulse extraction efficiency and has the potential to reduce the sensitivity of gain to temperature. The demonstration of direct diode pumping47 with Ho3+ opens up additional power-scaling opportunities.

Extending the laser wavelength further into the infrared has necessitated the use of fluoride glass fibre, owing to its low pho-non energy. Figure 4 shows the primary fluorescence transitions at around 3 μm from rare-earth cations doped into fluoride glass. The fluorescence spectra tend to cluster in a region that spans approxi-mately 0.6 μm. The most developed fibre laser in this spectral region is the Er3+-doped fluoride fibre laser, which can be tuned48 across the range of 2.71–2.88 μm. Engineering the stability and efficiency of the output from a fluoride fibre laser — a necessary requirement for commercial exploitation — is made possible by writing Bragg gratings49 and splicing glass caps50 to the ends of the fibre. The over-lap of the upper laser level absorption with highly developed diode lasers emitting at 0.98 μm, combined with effective cooling of Er3+-doped double-clad fluoride fibre, has resulted in output powers of up to 24 W at wavelengths of around 2.8 μm (ref. 24). When high-quality Er3+-doped double-clad fluoride fibre with a background loss of <100 dB km–1 is combined with a high Er3+ concentration and an optimally engineered fibre laser resonator, energy transfer upconver-sion (Fig. 5) effectively de-populates the lower laser level and ‘recy-cles’ the excitation to produce slope efficiencies of 35.6%51 — a value that exceeds the Stokes efficiency limit. Engineering the effective de-population of the lower laser level is required because the upper laser level has a shorter luminescence lifetime than the lower laser level. Recent studies52 suggest, however, that the energy transfer rate parameters may be at least an order of magnitude smaller in fibres

than in bulk glass, and thus the Stokes-exceeding slope efficiency in the above demonstration51 required a large Er3+ concentration. Cascade lasing53–55 the transitions of 2.8 μm and 1.5 μm offers a solu-tion to the problem of managing the heat generated due to the com-paratively low optical conversion efficiencies of single-transition Er3+ systems. Pump-excited state absorption from the upper laser level (Fig. 3) creates a roll-off in the calculated56 and measured51 output power and thus presents a significant problem for power-scaling the output from Er3+-doped fluoride fibre lasers pumped to the upper laser level. Ho3+-doped fluoride fibre57, which has a wide emission range of 2.8–3.02 μm, has the advantage of reduced pump-excited state absorption and a higher Stokes limit when the upper laser level is diode-pumped at 1.15 μm. However, the present maximum out-put power of this system is an order of magnitude lower than that of an Er3+-doped fluoride fibre laser because the low demand for diode lasers operating at 1.15 μm means that beam-combining and brightness-conserving power-scaling technologies have not been expanded to these pump sources. An opportunity to increase the power and wavelength of 3-μm-class fibre lasers has recently been demonstrated26 using cascade lasing of both the 2.9 μm and 2.1 μm transitions of Ho3+.

Host materialsThe typically long (>1 m) optical path length of a fibre laser means that glasses must exhibit low impurity, low scattering loss, a large Hruby parameter58 and a low maximum phonon energy. For emis-sion in the 1–2.2 μm region, the low loss and physical strength of silicate glasses has made them remarkably useful hosts for rare-earth cations emitting in this region. Fluoride glasses have high efficiency and moderate output power in the 2.3–3.5 μm region, and are now maturing to commercial levels. Beyond 3.5 μm, however, only a small number of glasses have the suitably low phonon energy that provides the necessarily high transmission of the fibre and sufficient radiative efficiency for the rare-earth cation transitions. Laser emis-sion at 3.9 μm was demonstrated under cryogenic conditions from a Ho3+-doped fluoride fibre laser29. This result remains the longest wavelength emitted from a fibre laser — no laser emission beyond 1 μm has yet been achieved using a rare-earth-cation-doped chalco-genide glass fibre.

0

2

4

6

8

Emission wavelength (µm)

Emis

sion

cro

ss-s

ectio

n (1

0–25 m

2 )

1.2 1.4 1.6 1.8 2.0 2.2 2.4 3.2 3.42.6 2.8 3.0

Oxide glass Fluoride glass

Er3+ (4I13/2 → 4I15/2)

Er3+ (4I11/2 → 4I13/2)Tm3+ (3F4 → 3H6)

Ho3+ (5I7 → 5I8)

Ho3+ (5I6 → 5I7)

Dy3+ (6H13/2 → 6H15/2)

Figure 4 | Fluorescence spectra of the rare-earth cation laser transitions used in fibre lasers emitting at wavelengths greater than 1.5 μm. The emission cross-section is shown as a function of emission wavelength. Included in the figure are the energy levels that participate in each of the laser transitions. The oxide glass refers to aluminosilicate and the fluoride glass refers to ZBLAN.




Silicate glasses remain the most successful fibre host materials. A high damage threshold59 of approximately 500 MW cm–2 for doped silicate glass ensures robust high-power operation, and fabrication using modified chemical vapour deposition ensures excellent glass purity. Silicate glasses have high melting points, relatively low ther-mal expansion coefficients, large tensile strengths, low refractive indices and low nonlinearities. All glasses have a low second-order nonlinearity, although loss to third-order nonlinearity processes such as stimulated Raman scattering can be substantial because of the long fibre length and large light intensity in the active core. Silicate glasses consist of strongly covalently bonded atoms that form a disordered matrix with a range of bond lengths and angles. Silica can sustain maximum phonon energies60 of up to 1,100 cm−1, which sets an upper limit on the emission wavelength. The giant covalent structure of silicate glasses leads to excellent mechanical properties. The maximum ‘lattice’ vibration or ‘clustron’ (phonon) energy is the most active in multiphonon decay, and the fluctuating Stark field sur-rounding the rare-earth cation arising from the vibration of the sur-rounding anions and cations induces non-radiative decay. Because the bandgap of silica is 9 eV, loss to two-photon absorption is small at infrared wavelengths and Rayleigh scattering loss from the fro-zen-in density fluctuations in the glass can be as low as 0.15 dB km–1 (ref. 61). When network formers such as Al2O3 and P2O5 are added to silica62, they create the solvation shells necessary for improved rare-earth cation solubility to counteract clustering. Unfortunately, Rayleigh scattering loss increases as a result of the extra refractive index inhomogeneities introduced by the addition of these dopants. Emission at 2 μm from Tm3+ and Ho3+ cations doped into silicate glasses is highly developed, and a number of commercial systems are now available. The longest laser wavelength from a silicate glass fibre

laser is currently 2.188 μm (ref. 63); lasing at longer wavelengths is not expected to be possible in the majority of silicate glasses.

One of the most successful compositions among fluoride glasses64 is ZBLAN65, which is comprised of 53 mol% ZrF4, 20 mol% BaF2, 4 mol% LaF3, 3 mol% AlF3 and 20 mol% NaF. The ZBLAN composi-tion can be varied to introduce certain characteristics: for example, adding PbF2, ZnF2 or CaF2 modifies the viscosity and refractive index; substituting a proportion of ZrF4 for ThF4 creates better stability against crystallization; and substituting ZrF4 with HfF4 regulates the glass stability while lowering the refractive index. Fluoride fibre pre-forms can be created by a number of casting processes66,67, although crystallization during cooling often creates scattering centres that limit the dimensions of the preform and the maximum length of use-ful fibre. Geometrical defects such as inclusions and bubbles formed during fabrication also cause scattering. Reducing the casting pres-sure of ZBLAN eliminates bubble formation to allow, when com-bined with anhydrous fluoride precursors, a minimum fibre loss of 0.65 dB km–1 at 2.59 μm and <20 dB km–1 at 2.9 μm (which is where the O–H impurity stretching vibration in ZBLAN is located68).

The maximum phonon energy of ZBLAN69 is approximately 565 cm–1, which allows fluorescence at room temperature to occur from rare-earth cation transitions comprised of energy gaps larger than 2,825 cm–1 (that is, λ < 3.5 μm). ZBLAN has low optical disper-sion, a low refractive index of 1.49 and a broad transmission window (defined here to have an attenuation of less than 200 dB km–1) in the range of 0.2–4.5 μm (ref. 70). Compared with silicate glasses, the lower maximum phonon energy of ZBLAN (and thus the extended infrared transmission) relates to the weaker bond strength and larger reduced mass of the atoms comprising the glass. The fluoride anion is singly charged, which, when combined with weaker bonding,

3H4

3H5

3H6Ion 1

3F4

Pump

Laser

Ion 2

Laser

4I9/2

4I11/2

4I13/2

4I15/2

Ion 1

Pump

Laser

Ion 2

Laser

a b

Figure 5 | Energy-transfer processes between neighbouring cations relevant to the functioning of Tm3+ and Er3+ fibre lasers. a, Cross-relaxation between Tm3+ cations creates two excited Tm3+ cations in the upper laser level for every pump photon excitation. b, Energy transfer between excited Er3+ cations forces the de-excitation of one Er3+ cation and the further excitation of the other Er3+ cation taking part in the process. Energy-transfer upconversion between two Tm3+ cations excited to the 3F4 level (that is, the reverse of cross-relaxation) can occur with a large rate parameter in a clustered system.




means a higher chemical reactivity compared with silicates, although rare-earth cations substitute the La3+ cations of the glass to provide comparatively higher doping levels without clustering. The optical damage threshold71 of ZBLAN for 10 ms pulses at 2.8 μm is approxi-mately 25 MW cm–2, which restricts the achievable peak power levels compared with silicates. However, because the Raman gain coeffi-cient is comparatively smaller72, significant loss to Raman scattering may never be a concern in future high-power infrared fibre lasers using this glass. The higher thermal loading resulting from the mod-erate efficiency of 3-μm-class fibre lasers (for which ZBLAN fibres are used), combined with the comparatively poorer physical proper-ties of fluoride glasses, has made power-scaling a significant chal-lenge. The maximum output power from ZBLAN-based fibre lasers is currently 24 W at 2.7 μm (ref. 24) and 20 W at 1.94 μm (ref. 33), although there are opportunities to increase this output power, as discussed below.

There is ongoing development in the fabrication and testing of glasses for achieving longer emission wavelengths. Germanate glasses73 have robust mechanical qualities, maximum phonon ener-gies74 of 900 cm–1 and large rare-earth cation solubilities, which allow cluster-reduced Tm3+ concentrations in fibres for high-efficiency 1.9 μm emission75 and short-length devices providing narrow-linewidth outout76. However, the practicality of germanate glasses for emission beyond 2 μm is reliant on effective removal of OH– impurity from the glass, which is difficult to achieve. Tellurite glass fibre77 is a promising alternative, but, like most oxide glasses fabricated from solid-state precursor materials, it contains an insufficiently low con-centration of OH–. However, energy transfer from excited Er3+ cati-ons to OH– impurities, combined with the low radiative efficiency of the upper laser level, is sufficient to suppress 3 μm lasing in state-of-the-art Er3+-doped tellurite glass78. Chalcogenides79 — glasses based on chalcogens S, Se or Te — are a widely investigated group whose high refractive indices result in large absorption and emission cross-sections. Maximum phonon energies depend on the exact composi-tion of the glass, with values of 350–425 cm–1 for sulphide glasses80, 250–300 cm–1 for selenide glasses81 and 150–200 cm–1 for telluride glasses82. Optical transmission reaches well into the mid-infrared — up to 15 μm in the case of telluride glass83 — and fluorescence has been measured in sulphide glass from a number of rare-earth cation transitions with emission wavelengths as long as 4.3 μm (ref. 84). The only demonstration of a fibre laser using a rare-earth-doped chalcogenide fibre was based on Nd3+ emitting at 1.08 μm (ref. 85), in which the addition of La2O3 to gallium chalcogen glass in order to avoid crystallization during fibre drawing86 most likely created energetic phonons that suppressed longer-wavelength emission87. Ge–Sb and Ge–As chalcogen glasses, for example, are strongly cova-lently bonded, which makes the incorporation of rare-earth cations a significant challenge. Alternative co-dopants for chalcogen-based glasses are being developed to improve rare-earth cation solubility and eliminate crystallization88, with the aim of supplying rare-earth-cation-doped chalcogenide fibres that will underpin future mid-infrared emission from a fibre laser source.

Pulsed fibre lasersFibre lasers cover a number of pulse width regimes, but the small dimensions of the active core invoke surface damage-threshold limitations. So far, silicate glass fibre has provided the highest peak power, largest pulse energy and widest variety of pulse widths from fibre lasers operating at 2 μm, primarily because of its large damage threshold and the overall maturity of 2 μm fibre lasers. Fibre lasers using Tm3+ and Ho3+ cations have been mode-locked for ultrashort pulse generation using nonlinear polarization rotation89, semicon-ductor saturable absorbers90 and carbon nanotubes91. Because the saturation energy is proportional to the core area, the increased mode size relevant to 2 μm fibre lasers compared with 1 μm fibre lasers

benefits energy extraction. Today’s shortest pulsewidth at 2 μm from a fibre is 108 fs (ref. 92). Mode-locked fibre lasers operating at 2 μm typically create solitons because silicate fibres are anomalously dis-persive at these wavelengths. Larger pulse energies can be achieved by engineering the overall dispersion using grating telescopes93 or highly Ge-doped fibre for dispersion compensation. Chirped pulse amplification of ultrashort pulses using Tm3+-doped silicate glass fibre leads to near-megawatt peak powers after recompression94. For applications that require longer pulses, active Q-switching95 is capa-ble of generating pulses measuring just a few tens of nanoseconds in duration96. Gain-switching97 using pulsed diode lasers emitting at 1.5 μm for direct excitation of the upper laser level can generate pulses of less than 2 ns in duration98.

Pulsed fibre laser sources emitting at longer wavelengths make use of fluoride glass, which has a lower surface optical damage thresh-old but potentially larger mode area than silicate glass. Although fluoride glass also benefits from a lower Raman gain coefficient72, it suffers from relatively lower optical efficiencies and weak thermo-mechanical properties. Nevertheless, steady progress has been made with recent demonstrations of Q-switched99 (90 ns pulsewidth and 0.9 kW peak power) and gain-switched100 (307 ns pulsewidth and 68 W peak power) operation of Er3+-doped ZBLAN fibre lasers. Recent demonstrations of Q-switched single-transition101 and cas-caded102 Ho3+-doped fluoride fibre lasers provide 70 ns pulses, emis-sion at 2.87 μm and two-wavelength output. Given that additional mode-size enhancement can be achieved by capping the ends of the fibre50, and that 5 kW peak power has already been demonstrated using fluoride fibre103, the output performance of pulsed fluoride fibre lasers is likely to see further improvement. The ability to gener-ate clean pulses with smooth transverse mode profiles that prevent localized large intensities will be particularly relevant to future high-peak-power fluoride fibre lasers.

ApplicationsMany applications require intense laser radiation in the near- or mid-infrared. In soft-tissue medicine, direct application of the Tm3+-doped silicate fibre laser has been widespread owing both to its commercial availability and to the fact that its emission wavelength overlaps with the absorption wavelength of the combination vibra-tion of the O–H bond in water at 1.94 μm (ref. 104). In the field of urology, Tm3+-doped silicate fibre lasers have been used to ablate and incise urinary tissues105,106, vapourize and resect the prostrate107,108 and manage urinary tract tumours, strictures and calculi109. Organs such as the kidneys110, brain111, skin112 and larynx113,114 all benefit from the low carbonization and excellent ablative properties115 of intense 2 μm light generated from the Tm3+-doped silicate fibre laser. The geometry of this fibre is well-suited to natural-orifice endoscopic procedures116, and the high efficiency and compact size of these devices suits all sur-gical environments. Moving to wavelengths closer to the fundamen-tal stretching vibration of O-H at 2.95 μm (ref. 104) has significant benefits. Almost no carbonization and comparatively faster ablation rates have been measured117 using moderate output power 2.7 μm emission from an Er3+-doped ZBLAN fibre laser, although Er3+- or Ho3+-doped fluoride lasers must become commercially available before extensive biomedical experiments on soft bodily tissues can be carried out. These lasers in particular have the potential to make a significant impact in medicine. Experiments involving free-electron lasers have also shown that lipid-rich tissues118, bone119 and protein-containing tissues120 can be substantially ablated with minimal col-lateral damage by using laser sources operating in the mid-infrared.

Military applications require intense light for directed energy applications and propagation over long distances within atmos-pheric transmission windows. Fibre lasers emitting at 1 μm with out-put powers of up to 50 kW have been used to detonate land mines, unexploded bombs and short-range rockets, all of which require a




near-diffraction-limited beam. Scattered 2 μm light from the Tm3+-doped silicate fibre laser cannot propagate through to the retina and is therefore much less of a hazard than lasers operating at shorter wavelengths. Future multi-kilowatt Tm3+-doped silicate fibre lasers operating in a single transverse mode could be used on a number of combat platforms for defence against cruise missiles, rockets and unmanned reconnaissance air vehicles. For infrared-guided missile countermeasures, the 3–5 μm atmospheric transmission window can be accessed using optical parametric oscillators containing zinc germanium phosphide121 or orientation-patterned GaAs122 pumped with Q-switched Ho3+-doped crystalline lasers that are resonantly pumped with Tm3+-doped silicate fibre lasers. Low-quantum-defect Tm3+-doped silicate fibre laser excitation of Ho3+ can be used to exploit the long lifetime of the 5I7 level in Ho3+. Zinc germanium phosphide optical parametric oscillators can be directly pumped with gain-switched Tm3+-doped silicate fibre lasers123 or a supercon-tinuum generated in the range of 1.9–4.5 μm using a combination of Tm3+-doped amplifying fibre and passive ZBLAN fibre124. The atmos-pheric window of 2.025–2.25 μm can be accessed directly using the output from Tm3+-doped silicate fibre lasers or Ho3+-doped silicate fibre lasers for remote atmospheric monitoring.

Mode-locked Tm3+-doped silicate fibre lasers have been used to generate octave-spanning supercontinua and carrier–envelope-offset frequency detection125. The mid-infrared contains molecular ‘fingerprints’ relating to strong rotational–vibrational absorption of molecular gases, liquids and solids such as greenhouse gases, pollut-ants and pharmaceuticals. Atmospheric components such as CO2, CO and NO2 have strong absorption lines at 2.8 μm, 2.4 μm and 2.9 μm, respectively, which can be accessed using differential absorp-tion LIDAR (light detection and ranging) employing fibre lasers. Various hydrocarbons, hydrochlorides and commonly used solvents display strong absorption features in the range of 3.2–3.6 μm; fibre lasers could offer a robust and efficient approach to the detection and real-time spectroscopy of these compounds. The ability to detect narcotics and explosives for security scanning as well as exhaust gases from industry and aircraft are all important applications that are benefiting from the development of reliable sources of mid-infra-red radiation.

Future directionsOver the past 20 years, the maximum output power available from a diode-pumped silicate fibre laser emitting at around 2 μm has increased by three orders of magnitude. However, questions regard-ing the long-term reliability of future multi-kilowatt-level 2 μm systems still remain. Blue light fluorescence from the 1G4 level in Tm3+-doped silicate fibre lasers due to pump-excited state absorp-tion at 0.79 μm (Fig. 3) could lead to photodarkening, and the heat generated from 30–40% of the absorbed pump power in today’s sys-tems could be difficult to manage. Access to flexible core designs is limited because of the large refractive index contrast between the active core and the pump core, which results from the large Tm3+ concentrations required for cross-relaxation. To combat these issues, resonantly exciting45 Ho3+-doped silicate fibre lasers with the output from multi-transverse-mode Tm3+-doped silicate fibre lasers is a promising solution. Exploiting the small quantum defect, all-silicate-glass construction and the negligible energy-transfer processes between Ho3+ cations can be accessed by using all-glass double-clad low-concentration (<1 mol%) Ho3+-doped silicate fibre pumped with high-power Tm3+-doped silicate fibre lasers using beam com-biners. Recent experiments on a prototype of this arrangement are certainly encouraging23. Alternatively, pumping low-concentration (<2.5 mol%) Tm3+-doped fibre with diode lasers operating at 1.53–1.56 μm in a similar way to Er3+-doped silicate glass fibre lasers126 will lead to improvements in efficiency127, and transferring fibre tech-nologies developed for the near-infrared to 2 μm wavelengths128 will

provide opportunities to exploit this important region of the near-infrared spectrum even further.

The development of silicate glass optical fibres with moderately low concentrations of rare-earth cations holds great promise for pro-viding high-power emission at 2 μm; the same may also be true for fluoride fibre lasers. Thermally induced performance degradation in today’s fluoride fibre lasers is consistently reached before the onset of nonlinear loss. Reducing the rare-earth cation concentration in the core of a cascaded fibre laser reduces the absorption coefficient and thus the heat load. Employing fibre geometries that have appropri-ately designed ratios between the active- and pump-core areas can effectively remove the heat, thus lowering the core temperatures and significantly improving the power-scaling potential. Lower concen-trations of rare-earth cations in fluoride fibres will allow the use of more stable and more flexible ZBLAN compositions that have lower propensities for crystallization and defect formation. Joining diode lasers with fluoride fibre to create fluoride glass pump combiners will provide distributed pumping opportunities for fluoride fibre lasers, and aperture-scaling the output using spectral beam combining129 will relieve the power requirements of individual fibre lasers.

The realization of fibre lasers capable of emitting beyond 3 μm is a significant challenge. In the past, host material issues such as low purity, large maximum phonon energies and low optical dam-age thresholds have impeded progress, although there are a num-ber of opportunities for carrying the benefits of fibre laser emission well into the mid-infrared. Current studies into the development of efficient long-wavelength pump sources are combating the quantum defect problem, and recent cascaded fibre laser experiments26 using Ho3+ have demonstrated mid-infrared emission at room tempera-ture using diode pumping. A number of rare-earth cation transitions are capable of generating mid-infrared fluorescence in hosts that can be drawn into fibre; for example, chalcogenide glass130 and silver hal-ide crystals131,132. We must now find ways to incorporate rare-earth cations into chalcogenide glasses without devitrification or increas-ing the phonon energy, particularly because state-of-the-art chalco-genide fibre has the necessary low loss to support moderate output power levels. Utilizing the large Raman gain coefficient of chalcoge-nide glass in a Raman fibre laser arrangement has strong potential given the increasing availability of commercial low-loss chalcoge-nide fibre, and gas-filled hollow core fibres could offer a versatile solution as Raman shifters133 for extending the available wavelength range. The development of fibre lasers emitting in the mid-infrared is a growing field of photonics that has already demonstrated many successes, but still faces a number of significant challenges.

References1. Snitzer, E. et al. Double clad, offset core Nd fiber laser in Optical Fiber Sensors,

Vol. 2, paper PD5 of OSA Technical Digest Series (OSA, 1988).2. Zenteno, L. High-power double clad fiber lasers. J. Lightwave Technol. 11,

1435–1446 (1993).3. Jeong, Y., Sahu, J. K., Payne, D. N. & Nilsson, J. Ytterbium-doped large-core

fiber laser with 1.36 kW continuous-wave output power. Opt. Express 12, 6088–6092 (2004).

4. Poole, S. B., Payne, D. N. & Fermann M. E. Fabrication of low-loss optical fibers containing rare earth ions. Electron. Lett. 21, 737–738 (1985).

5. Faist, J. et al. Quantum cascade laser. Science 264, 553–556 (1994).6. Vurgaftman, I. & Meyer, J. R. Analysis of limitations to wallplug efficiency and

output power for quantum cascade lasers. J. Appl. Phys. 99, 123108 (2006).7. Deloach, L. D. et al. Transition metal-doped zinc chalcogenides: spectroscopy

and laser demonstration of a new class of gain media. IEEE J. Quant. Electron. 32, 885–895 (1996).

8. Sanghera, J. et al. Ceramic laser materials. Materials 5, 258–277 (2012).9. http://www.ipgphotonics.com/.10. Sorokin, E., Naumov, S. & Sorokina, I. T. Ultrabroadband infrared solid-state

lasers. IEEE J. Sel. Top. Quant. Electron. 11, 690–712 (2005).11. Garbuzov, D. Z. et al. 2.3–2.7 μm room temperature CW operation of

InGaAsSb–AlGaAsSb broad waveguide SCH-QW diode lasers. IEEE Photon. Tech. Lett. 11, 794–796 (1999).




12. Vodopyanov, K. L. Mid-infrared optical parametric generator with extra-wide (3–19 μm) tunability: applications for spectroscopy of two dimensional electrons in quantum wells. J. Opt. Soc. Am. B 16, 1579–1586 (1999).

13. Shen, Y. et al. PPMgLN based high power optical parametric oscillator pumped by Yb3+-doped fiber amplifier incorporating active pulse shaping. IEEE J. Sel. Top. Quant. Electron. 15, 385–392 (2009).

14. Pask, H. M. et al. Ytterbium-doped silica fiber lasers — versatile sources for the 1–12 μm region. IEEE J. Sel. Top. Quant. Electron. 1, 2–13 (1995).

15. Offerhaus, H. L. et al. High-energy single-transverse-mode Q-switched fiber laser based on a multimode large-mode-area erbium-doped fiber. Opt. Lett. 23, 1683–1685 (1998).

16. Knight, J. C. et al. Large mode area photonic crystal fibre. Electron. Lett. 34, 1347–1348 (1998).

17. Rottwitt, K. et al. Scaling of the Raman gain coefficient: applications to germanosilicate fibers. J. Lightwave Technol. 21, 1652–1662 (2003).

18. Koponen, J. J., Söderlund, M. J., Hoffman, H. J. & Tammela, S. K. T. Measuring photodarkening from single-mode ytterbium doped silica fibers. Opt. Express 14, 11539–11544 (2006).

19. Jetschke, S., Unger, S., Röpke, U. & Kirchhof, J. Photodarkening in Yb doped fibers: experimental evidence of equilibrium states depending on the pump power. Opt. Express 15, 14838–14843 (2007).

20. Fomin, V. et al. 10 kW single-mode fiber laser, presented at the 5th International Symposium on High-Power Fiber Lasers and Their Applications (2010).

21. Jeong, Y. et al. Erbium:ytterbium codoped large-core fiber laser with 297 W continuous-wave output power. IEEE J. Sel. Top. Quant. Electron. 13, 573–579 (2007).

22. Ehrenreich, T. et al. 1 kW, all-glass Tm:fiber laser. SPIE Photonics West 2010: LASE, Fibre Lasers VII: Technology, Systems and Applications, Conference 7850 (2010).

23. Hemming, A. V. et al. Development of resonantly cladding-pumped holmium-doped fibre lasers. Proc. SPIE 8237, 82371J (2012).

24. Tokita, S. et al. Liquid-cooled 24 W mid-infrared Er:ZBLAN fiber laser. Opt. Lett. 34, 3062–3064 (2009).

25. Jackson, S. D. Single-transverse-mode 2.5-W holmium-doped fluoride fiber laser operating at 2.86 μm. Opt. Lett. 29, 334–336 (2004).

26. Li, J., Hudson, D. D. & Jackson S. D. High-power diode-pumped fiber laser operating at 3 μm. Opt. Lett. 36, 3642–3644 (2011).

27. Carbonnier, C., Többen, H., Unrau, U. B. Room temperature CW fibre laser at 3.22 μm. Electron. Lett. 34, 893–894 (1998).

28. Tobben, H. Room temperature CW fibre laser at 3.5 μm in Er3+-doped ZBLAN glass. Electron. Lett. 28, 1361–1363 (1992).

29. Schneider, J., Carbonnier, C. & Unrau, U. B. Characterization of a Ho3+-doped fluoride fiber laser with a 3.9 μm emission wavelength. Appl. Opt. 36, 8595–8600 (1997).

30. Brown, D. C. & Hoffman, H. J. Thermal, stress, and thermo-optic effects in high average power double-clad silica fiber lasers. IEEE J. Quant. Electron. 37, 207–217 (2001).

31. Esterowitz, L., Eckardt, R. C., Allen, R. E. Long-wavelength stimulated-emission via cascade laser action in Ho-YLF. Appl. Phys. Lett. 35, 236–239 (1979).

32. International Standard ISO 20473 “Optics and Photonics — Spectral bands” (2007).

33. Eichhorn, M. & Jackson, S. D. Comparative study of continuous wave Tm3+-doped silica and fluoride fiber lasers. Appl. Phys. B 90, 35–41 (2008).

34. Jackson, S. D. et al. High-power 83 W holmium-doped silica fiber laser operating with high beam quality. Opt. Lett. 32, 241–243 (2007).

35. El-Agmy, R. M. & Al-Hosiny, N. M. 2.31 μm laser under up-conversion pumping at 1.064 μm in Tm3+:ZBLAN fibre lasers. Electron. Lett. 46, 936–937 (2010).

36. Jackson, S. D. Continuous wave 2.9 μm dysprosium-doped fluoride fiber laser. Appl. Phys. Lett. 83, 1316–1318 (2003).

37. Greaves, G. N. EXAFS and the structure of glass. J. Non-cryst. Solids 71, 203–217 (1985).

38. Caro, P., Beaury, O. & Antic, E. Nephelauxetic effect for 4fN configurations in solid-state. J. Phys.-Paris 37, 671–676 (1976).

39. Hanna, D. C., Percival, R. M., Smart, R. G. & Tropper A. C. Efficient and tunable operation of a Tm-doped fibre laser. Opt. Commun. 75, 283–286 (1990).

40. Jackson, S. D. & King, T. A. High-power diode-cladding-pumped Tm-doped silica fiber laser. Opt. Lett. 23, 1462–1464 (1998).

41. Hayward, R. A. et al. Efficient cladding-pumped Tm-doped silica fibre laser with high power singlemode output at 2 μm. Electron. Lett. 36, 711–712 (2000).

42. Moulton, P. F. et al. Tm-doped fiber lasers: fundamentals and power scaling. IEEE J. Sel. Top. Quant. Electron. 15, 85–92 (2009).

43. Clarkson, W. A. et al. High-power cladding-pumped Tm-doped silica fiber laser with wavelength tuning from 1860 to 2090 nm. Opt. Lett. 27, 1989–1991 (2002).

44. Jackson, S. D. Cross relaxation and energy transfer upconversion processes relevant to the functioning of 2 μm Tm3+-doped silica fibre lasers. Opt. Commun. 230, 197–203 (2004).

45. Jackson, S. D. Midinfrared holmium fiber lasers. IEEE J. Quant. Electron. 42, 187–191 (2006).

46. Oh, K. et al. Continuous-wave oscillation of thulium-sensitized holmium-doped silica fiber laser. Opt. Lett. 19, 278–280 (1994).

47. Jackson, S. D., Bugge, F. & Erbert, G. Directly diode-pumped holmium fiber lasers. Opt. Lett. 32, 2496–2498 (2007).

48. Tokita, S. et al. Stable 10 W Er:ZBLAN fiber laser operating at 2.71–2.88 μm. Opt. Lett. 35, 3943–3945 (2010).

49. Faucher, D. et al. 20 W passively cooled single-mode all-fiber laser at 2.8 μm. Opt. Lett. 36, 1104–1106 (2011).

50. Bernier, M. et al. Bragg gratings photoinduced in ZBLAN fibers by femtosecond pulses at 800 nm. Opt. Lett. 32, 454–456 (2007).

51. Faucher, D., Bernier, M., Caron, N. & Vallee, R. Erbium-doped all-fiber laser at 2.94 μm. Opt. Lett. 34, 3313–3315 (2009).

52. Gorjan, M., Marincek, M. & Copic, M. Role of interionic processes in the efficiency and operation of erbium-doped fluoride fiber lasers. IEEE J. Quant. Electron. 47, 262–273 (2011).

53. Schneider, J. Mid-infrared fluoride fiber lasers in multiple cascade operation. IEEE Photon. Tech. Lett. 7, 354–356 (1995).

54. Pollnau, M. et al. Three-transition cascade erbium laser at 1.7, 2.7, and 1.6 μm. Opt. Lett. 22, 612–614 (1997).

55. Jackson, S. D. High-power erbium cascade fibre laser. Electron. Lett. 45, 830–832 (2009).

56. Li, J. & Jackson, S. D. Numerical modeling and optimization of diode pumped heavily-erbium-doped fluoride fiber lasers. IEEE J. Quant. Electron. 48, 454–464 (2012).

57. Wetenkamp, L., West, G. F. & Tobben. H. Optical properties of rare earth-doped ZBLAN glasses. J. Non-cryst. Solids 140, 35–40 (1992).

58. Hruby, A. Evaluation of glass-forming tendency by means of DTA. Czech J. Phys. B22, 1187–1193 (1972).

59. Tünnermann, A. et al. The renaissance and bright future of fibre lasers. J. Phys. B 38, S681–S693 (2005).

60. Layne, C. B., Lowdermilk, W. H. & Weber, M. J. Multiphonon relaxation of rare-earth ions in oxide glasses. Phys. Rev. B 16, 10–20 (1977).

61. Saito, K. et al. Limit of the Rayleigh scattering loss in silica fiber. Appl. Phys. Lett. 83, 5175–5177 (2003).

62. Arai, K. et al. Aluminium or phosphorous co-doping effects on the fluorescence properties of neodymium-doped silica glass. J. Appl. Phys. 59, 3430–3436 (1986).

63. Sacks, Z. S., Schiffer, Z. & David, D. Long wavelength operation of double-clad Tm:silica fiber lasers. Proc. SPIE 6453, 645320 (2007).

64. Zhu, X. & Peyghambarian, N. High-power ZBLAN glass fiber lasers: review and prospect. Adv. Optoelectron. 2010, 501956 (2010).

65. Ohsawa, K., Shibata, T., Nakamura, K. & Yoshida, S. Fluorozirconate glasses for infrared transmitting optical fibers. Paper 1.1, Technical Digest, 7th European Conference on Optical Communication (1981).

66. Mitachi, S., Miyashita, T. & Kanamori, T. Fluoride-glass cladded optical fibers for mid-infra-red ray transmission. Electron. Lett. 17, 591–592 (1981).

67. Tran, D. C., Fisher, C. F. & Sigel, G. H. Jr. Fluoride glass performs prepared by a rotation casting process. Electron. Lett. 18, 657–658 (1982).

68. Carter, S. F. et al. Low loss fluoride fibre by reduced pressure casting. Electron. Lett. 26, 2115–2117 (1990).

69. Almeida, R. M. & Mackenzie, J. D. Vibrational spectra and structure of fluorozirconate glasses. J. Chem. Phys. 74, 5954–5961 (1981).

70. Day, C. R. et al. Fluoride fibres for optical transmission. Opt. Quant. Electron. 22, 259–277 (1990).

71. Zhu, X. & Jain, R. 10-W-level diode-pumped compact 2.78 μm ZBLAN fiber laser. Opt. Lett. 32, 26–28 (2007).

72. Fortin, V., Bernier, M., Carrier, J. & Vallée, R. Fluoride glass Raman fiber laser at 2185 nm. Opt. Lett. 36, 4152–4154 (2011).

73. Wang, J., et al. Fabrication and optical-properties of lead-germanate glasses and a new class of optical fiber lasers doped with Tm3+, J. Appl. Phys. 73, 8066–8075 (1993).

74. Lincoln J. R. et al. New class of fibre laser based on lead-germanate glass. Electron. Lett. 28, 1021–1022 (1992).

75. Wu, J., Yao, Z., Zong, J. & Jiang, S. Highly efficient high-power thulium-doped germanate glass fiber laser. Opt. Lett. 32, 638–640 (2007).

76. Geng, J., Wu, J., Jiang, S. & Yu, J. Efficient operation of diode-pumped single-frequency thulium-doped fiber lasers near 2 μm. Opt. Lett. 32, 355–357 (2007).




77. Mori, A., Ohishi, Y. & Sudo, S. Erbium-doped tellurite glass fibre laser and amplifier. Electron. Lett. 33, 863–864 (1997).

78. Gomes, L. et al. Energy level decay and excited state absorption processes in erbium-doped tellurite glass. J. Appl. Phys. 110, 083111 (2011).

79. Eggleton, B. J., Luther-Davies, B. & Richardson, K. Chalcogenide photonics. Nature Photon. 5, 141–148 (2011).

80. Julien, C. et al. Raman and infrared spectroscopic studies of Ge–Ga–Ag sulfide glasses. Mat. Sci. Eng. B22, 191–200 (1994).

81. Weszka, J. et al. Raman scattering in In2Se3 and InSe2 amorphous films. J. Non-cryst. Solids 265, 98–104 (2000).

82. Uemura, O., Hayasaka, N., Tokairin, S. & Usuki, T. Local atomic arrangement in Ge–Te and Ge–S–Te glasses. J. Non-cryst. Solids 205, 189–193 (1996).

83. Maurugeon, S. et al. Telluride glass step index fiber for the far infrared. J. Lightwave Technol. 28, 3358–3363 (2010).

84. Schweizer, T., Hewak, D. W., Samson, B. N., Payne, D. N. Spectroscopic data of the 1.8-, 2.9- and 4.3-μm transitions in dysprosium-doped gallium lanthanum sulfide glass. Opt. Lett. 21, 1594–1596 (1996).

85. Schweizer, T. et al. Rare-earth doped chalcogenide glass laser. Electron. Lett. 32, 666–667 (1996).

86. Brady, D. J. & Schweizer, T. Minimum loss predictions and measurements in gallium lanthanum sulfide based glasses and fibers. J. Non-cryst. Solids 242, 92–98 (1998).

87. Seddon, A. B. et al. Progress in rare-earth-doped mid-infrared fiber lasers. Opt. Express 18, 26704–26719 (2010).

88. Aggarwal, I. D. & Sanghera, J. S. Development and applications of chalcogenide glass optical fibers at NRL. J. Optoelectron. Adv. Mat. 4, 665–678 (2002).

89. Nelson, L. E., Ippen, E. P. & Haus, H. A. Broadly tunable sub-500 fs pulses from an additive-pulse mode-locked thulium-doped fiber ring laser. Appl. Phys. Lett. 67, 19–21 (1995).

90. Sharp, R. C., Spock, D. E., Pan, N. & Elliot, J. 190-fs passively mode-locked thulium fiber laser with a low threshold. Opt. Lett. 21, 881–883 (1996).

91. Solodyankn, M. A. et al. Mode-locked 1.93 μm thulium laser with a carbon nanotube absorber. Opt. Lett. 33, 1336–1338 (2008).

92. Imeshev, I. & Fermann, M. E. 230-kW peak power femtosecond pulses from a high power tunable source based on amplification in Tm-doped fiber. Opt. Express 13, 7424–7431 (2005).

93. Engelbrecht, M., Haxsen, F., Ruehl, A., Wandt, D. & Kracht, D. Ultrafast thulium-doped fiber-oscillator with pulse energy of 4.3 nJ. Opt. Lett. 33, 690–692 (2008).

94. Haxsen, F. et al. Pulse energy of 151-nJ from ultrafast thulium-doped chirped-pulse fiber amplifier. Opt. Lett. 35, 2991–2993 (2010).

95. El-Sherif, A. F. & King, T. A. High-peak-power operation of a Q-switched Tm3+-doped silica fiber laser operating near 2 μm. Opt. Lett. 28, 22–24 (2003).

96. Eichhorn, M. & Jackson, S. D. High-pulse-energy actively Q-switched Tm3+-doped silica 2 μm fiber laser pumped at 792 nm. Opt. Lett. 32, 2780–2782 (2007).

97. Jiang, M. & Tayebati, P. Stable 10 ns, kilowatt peak-power pulse generation from a gain-switched Tm-doped fiber laser. Opt. Lett. 32, 1797–1799 (2007).

98. Ding, J. W. et al. A monolithic thulium doped single mode fiber laser with 1.5 ns pulsewidth and 8 kW peak power. Proc. SPIE 7914, 79140X (2011).

99. Tokita, S. et al. 12 W Q-switched Er:ZBLAN fiber laser at 2.8 μm. Opt. Lett. 36, 2812–2814 (2011).

100. Gorjan, M., Petkovšek, R., Marinček, M. & Čopič, M. High-power pulsed diode-pumped Er:ZBLAN fiber laser. Opt. Lett. 36, 1923–1925 (2011).

101. Hu, T., Hudson, D. D. & Jackson, S. D. Actively Q-switched 2.9 μm Ho3+Pr3+-doped fluoride fiber laser. Opt. Lett. 37, 2145–2147 (2012).

102. Li, J., Hu, T. & Jackson, S. D. Q-switched fiber cascade laser. Opt. Lett. 37, 2208–2210 (2012).

103. Eichhorn, M. High-peak-power Tm-doped double-clad fluoride fiber amplifier. Opt. Lett. 30, 3329–3331 (2005).

104. Hale, G. M. & Querry, M. R. Optical-constants of water in 200-nm to 200-μm wavelength region. Appl. Opt. 12, 555–563 (1973).

105. Fried, N. M. & Murray, K. E. New technologies in enodurology — high power thulium fiber laser ablation of urinary tissues at 1.94 μm. J. Endourol. 19, 25–31 (2005).

106. Bach, T., Herrman, T. R., Cellarius, C. & Gross, A. J. Bladder neck incision using a 70 W, 2 micron continuous wave laser (RevoLix). World J. Urol. 25, 263–267 (2007).

107. Fried, N. M. High-power laser vaporization of the canine prostrate using a 110 W thulium fiber laser at 1.91 μm. Las. Surg. Med. 36, 52–56 (2005).

108. Xia, S. J. et al. Thulium fiber laser versus standard transurethral resection of the prostrate: a randomized prospective trial. Eur. Urol. 53, 382–390 (2008).

109. Kourambas, J., Delvecchio, F. C. & Preminger, G. M. Low-power holmium laser for the management of urinary tract calculi, strictures and tumors. J. Endourol. 15, 529–532 (2001).

110. Bui, M. H. et al. Less smoke and minimal tissue carbonization using a thulium laser for laparoscopic partial nephrectomy without hilar clamping in a porcine model. J. Endourol. 21, 1107–1111 (2007).

111. Bilici, T. et al. Development of a thulium (Tm:YAP) laser system for brain tissue ablation. Las. Med. Sci. 26, 699–706 (2011).

112. Chen, B. et al. Histological and modeling study of skin thermal injury to 2 μm laser radiation. Las. Surg. Med. 40, 358–370 (2008).

113. Zeitels, S. M. et al. Office-based and microlaryngeal applications of fiber-based thulium laser. Ann. Otol. Rhinol. Laryngol. 115, 891–896 (2006).

114. Burns, J. A. et al. Thermal damage during thulium laser dissection of laryngeal soft tissue is reduced with air cooling: ex vivo calf model study. Ann. Otol. Rhinol. Laryngol. 116, 853–857 (2007).

115. Pierce, M. C., Jackson S. D., Dickinson, M. R. & King T. A. Laser-tissue interaction with a high-power 2 μm fiber laser: preliminary studies with soft tissue. Las. Surg. Med. 25, 407–413 (1999).

116. Hazey, J. W. et al. Natural-orifice transgastric endoscopic peritoneoscopy in humans: initial clinical trial. Surg. Endosc. 22, 16–20 (2008).

117. Pierce, M. C. et al. Laser-tissue interaction with a continuous wave 3 μm fiber laser: preliminary studies with soft tissue. Las. Surg. Med. 26, 491–495 (2000).

118. Anderson R. R. et al. Selective photothermolysis of lipid-rich tissues: a free electron study. Las. Surg. Med. 38, 913–919 (2006).

119. Peavy, G. M., Reinisch, L., Payne, J. T. & Venugopalan, V. Comparison of cortical bone ablations by using infrared laser wavelengths 2.9 to 9.2 μm. Las. Surg. Med. 26, 421–434 (1999).

120. Edwards, G. et al. Tissue ablation by a free-electron laser tuned to the amide II band. Nature 371, 416–419 (1994).

121. Lippert, E. et al. Midinfrared laser source with high power and beam quality. Appl. Opt. 45, 3839–3845 (2006).

122. Kieleck, C. et al. High-efficiency 20–50 kHz mid-infrared orientation-patterned GaAs optical parametric oscillator pumped by a 2 μm holmium laser. Opt. Lett. 34, 262–264 (2009).

123. Creeden, D. et al. Mid-infrared ZnGeP2 parametric oscillator directly pumped by a pulsed 2 μm Tm-doped fiber laser. Opt. Lett. 33, 315–317 (2008).

124. Kulkarni, O. P. et al. Supercontinuum generation from ~1.9 to 4.5 μm in ZBLAN fiber with high average power generation beyond 3.8 μm using a thulium-doped fiber amplifier. J. Opt. Soc. Am. B 28, 2486–2498 (2011).

125. Phillips, C. R. et al.Self-referenced frequency comb from a Tm-fiber amplifier via PPLN waveguide supercontinuum generation. Paper PDPA5 in CLEO:2011 — Laser Applications to Photonic Applications, OSA Technical Digest (OSA, 2011).

126. Zhang, J., Fromzel, V. & Dubinskii, M. Resonantly cladding-pumped Yb-free Er-doped LMA fiber laser with record high power and efficiency. Opt. Express 19, 5574–5578 (2011).

127. Meleshkevich, M. et al. 415W single-mode CW thulium fiber laser in all-fiber format. Paper CP2_3 in CLEO/Europe and IQEC 2007 Conference Digest (OSA, 2007).

128. Modsching, N. et al. Lasing in thulium-doped polarizing photonic crystal fiber. Opt. Lett. 36, 3873–3875 (2011).

129. Fan, T. Y. Laser beam combining for high-power, high-radiance sources. IEEE J. Sel. Top. Quant. Electron. 11, 567–577 (2005).

130. Shaw, L. B. et al. Mid-wave IR and long-wave IR laser potential of rare-earth doped chalcogenide glass fiber. IEEE J. Quant. Electron. 48, 1127–1137 (2001).

131. Nagli, L., Gayer, O. & Katzir, A. Middle-infrared luminescence of praseodymium ions in silver halide crystals and fibers. Opt. Lett. 30, 1831–1833 (2005).

132. Rave, E. et al. Silver halide photonic crystal fibers for the middle infrared. Appl. Opt. 43, 2236–2241 (2004).

133. Jones, A. M. et al. Mid-infrared gas filled photonic crystal fiber laser based on population inversion. Opt. Express 19, 2309–2316 (2011).

AcknowledgementsThe author acknowledges financial support from the Australian Research Council through receipt of a Queen Elizabeth II Fellowship and, the Discovery Projects and Centre of Excellence funding schemes.

