A promising mid-infrared transition metal laser ion

The mid-infrared (mid-IR) wavelength region (2–5μm) is used in a number of important applications such as environmental sensing, infrared countermeasures against heat-seeking missiles, and laser radar. It is also extensively used in medicine, spectroscopy, and manufacturing. These applications are all driving a pressing need for robust, compact, solid-state, tunable laser sources operating at room temperature in the mid-IR range. In this context, transition metal ions have always been of great interest from the very beginnings of laser development efforts because of their broadband emission. Recent developments in room temperature Cr²⁺ lasers have resulted in impressive performance reports, but still face development and power-scaling roadblocks.

Ruby (Cr³⁺:Al₂O₃), the first material to demonstrate laser action in 1960,¹ turned out to be the exception to the rule that transition metal lasers are broadly tunable. Tunable transition metal laser development started with flashlamp-pumped demonstrations of Ni²⁺- and Co²⁺-doped fluoride hosts in 1963.^2,3 However, laser tuning in these hosts is often less than the observed fluorescence of their active ions: this restriction in tuning range is now known to be caused by excited state absorption (ESA).⁴ In addition, these materials were found to operate poorly or not at all at room temperature due to nonradiative quenching of the excited state populations.

Tunable chromium ion lasing was only discovered with Cr³⁺:alexandrite in 1979.⁵ The wait was even longer for Cr⁴⁺ and Cr²⁺ infrared laser materials. In 1995, researchers at the Lawrence Livermore National Laboratory reported a new class of room-temperature, widely tunable, mid-IR laser materials that did not exhibit the ESA or nonradiative quenching problems of previous transition metal lasers.^6,7 The active ion was Cr²⁺ and the host materials consisted of II-VI semiconductor compounds. While the initial Livermore report used ZnS and ZnSe as host materials, other researchers, including my group at the Air Force Research Laboratory, showed that other II-VI materials such as CdSe,^8,9 and even ternary mixtures of Cd_xMn_1−xTe,¹⁰ could be used as Cr²⁺ laser host materials. However, due to its favorable material properties, ZnSe remains the most widely used host material. Doping of ZnSe is typically done via thermal diffusion of CrSe at temperatures just below the ZnSe melting point.

Cr²⁺:ZnSe lasers have several advantages including flexible pumping schemes, impressive output powers, wide tunability and room-temperature operation. The broad absorption band in the 1.5–2.1μm region allows their pumping from a wide variety of sources such as Er, Tm, and Ho solid-state lasers, Raman-shifted Nd lasers, Co²⁺ lasers, color center lasers, and InGaAsP semiconductor lasers. Optical-to-optical conversion efficiency in these lasers runs above 60% with output power, initially at the sub-Watt level, recently scaled up to 18.5W by Carrig et al.¹¹ The broadband emission of the Cr²⁺ ion in ZnSe has led to extensive tunability of laser output in the 2000–3100nm range using an astigmatically compensated x-folded cavity and a Lyot filter for wavelength selection with a linewidth less than 0.4nm.¹² The broadband emission of the Cr²⁺ ion and its similarity to Ti³⁺ also make it attractive for infrared ultrashort pulse generation. Passive modelocking was first demonstrated by Sorokina et al. with 4ps-pulse widths and average output power as high as 400mW.¹³ Recently, pulses shorter than 100fs with 75mW of output power were achieved using an InAs/GaSb saturable absorber.¹⁴

The strong thermal lensing of Cr²⁺-doped II-VI materials is a major issue when designing Cr²⁺ lasers and especially when scaling power. A thin disk design (see Figure 1) trades the benefit of reduced thermal lensing for the added complexity of multi-pass pumping to ensure the efficient absorption of pump power. However, it does not solve the problem of the overall temperature increase that gets worse as disk thickness gets smaller, leading to higher nonradiative relaxation rates.¹⁵ An interesting possibility would be an improved fiber format: a fiber with the proper core radius and core/cladding refractive indices could confine beam propagation such that it would become insensitive to radial temperature gradients. Unfortunately, expertise in silica fiber-pulling does not translate easily to producing fiber cores from polycrystalline II-VI materials. To date, promising ideas to fabricate Cr²⁺-doped II-VI fibers are still lacking.

In summary, the history of infrared transition metal lasers closely follows that of lasers themselves with significant advances following the advent of the Cr²⁺ laser. The 1995 innovation of pairing Cr²⁺ with a II-VI semiconductor host has led to the demonstration of broadly tunable, room-temperature, mid-IR lasers. Cr²⁺ lasers have achieved high output power, tuning ranges greater than 1000nm, and modelocked operation at pulse widths shorter than 100fsec. Thermal lensing, thermally induced emission quenching, and concentration quenching are issues that still challenge further power scaling. New ideas such as direct electrical pumping and fabrication of a Cr²⁺ fiber laser are currently being developed. To conclude, the use of Cr²⁺ lasers in future applications looks very promising.