Due to strong water absorption and high transmittance in the atmosphere, mid-infrared (MIR) lasers operating at the 2.7 µm wavelength band have attracted increasing attention and play a significant role in applications including medical, biological, remote sensing, free-space communication, etc.^[1–3]. It can also be used as the pump source for an optical parametric oscillator (OPO) and an optical parametric amplifier (OPA) to generate even longer wavelength MIR lasers^[4,5]. ${\mathrm{Er}}^{3+}$ is a well-known ion with MIR transition around 2.7 µm (${{}^4\mathrm{I}}_{11/2}\to {{}^4\mathrm{I}}_{13/2}$). Compared with ${\mathrm{Ho}}^{3+}$ ions (${{}^5\mathrm{I}}_6\to {{}^5\mathrm{I}}_7$ transition around 2.9 µm), ${\mathrm{Er}}^{3+}$ ions have been extensively studied for several reasons. First, the branching ratio of ${\mathrm{Er}}^{3+}$ is twice that of ${\mathrm{Ho}}^{3+}$ for 3.0 µm laser transition. Second, ${\mathrm{Er}}^{3+}$-doped crystals can solve the so-called “bottleneck” at 3.0 µm laser transition by high doping concentration and the efficient Auger up-conversion process, which makes it easier to realize the population inversion than ${\mathrm{Ho}}^{3+}$-doped crystals. Third, ${\mathrm{Er}}^{3+}$-doped crystals can be pumped by commercial 970 nm laser diodes (LDs), but ${\mathrm{Ho}}^{3+}$-doped crystals must be pumped by the lasers around 1150 nm, which are mainly based on OPOs and Raman lasers^[6–8].

However, it is worth noting that ${\mathrm{Er}}^{3+}$-doped crystals face a serious “bottleneck” at the 2.7 µm laser transition, where the lifetime of the upper energy level (${{}^4\mathrm{I}}_{11/2}$) is much shorter than that of the lower energy level (${{}^4\mathrm{I}}_{13/2}$), resulting in the self-terminating effect^[9–13]. In view of this problem, there are two mainstream solutions. One solution is to increase the doping concentration of ${\mathrm{Er}}^{3+}$ ions (atomic fraction of >30%) to increase the cross-relaxation $[{{}^4\mathrm{S}}_{3/2}({{}^2\mathrm{H}}_{11/2})\to {{}^4\mathrm{I}}_{15/2}]+({{}^4\mathrm{I}}_{15/2}\to {{}^4\mathrm{I}}_{13/2})+({{}^4\mathrm{I}}_{15/2}\to {{}^4\mathrm{I}}_{9/2}\to {{}^4\mathrm{I}}_{11/2})$ and up-conversion ${(}^4{\mathrm{I}}_{13/2}\to {{}^4\mathrm{I}}_{15/2})+({{}^4\mathrm{I}}_{13/2}\to {{}^4\mathrm{I}}_{9/2}\to {{}^4\mathrm{I}}_{11/2})$ processes to effectively decrease the lifetime of ${{}^4\mathrm{I}}_{13/2}$. Nevertheless, on one hand, high ${\mathrm{Er}}^{3+}$ doping concentration can also lead to a large reduction in the lifetime of upper energy level ${{}^4\mathrm{I}}_{11/2}$; on the other hand, the thermal conductivity can also be reduced at the high doping level, resulting in the serious thermal effect. The above two factors impede their applications in high-power, high-energy, and high-beam-quality MIR lasers. Another method does not require a high ${\mathrm{Er}}^{3+}$ doping concentration, in which appropriate deactivated ions are introduced to attenuate the lifetime of the lower energy level (${{}^4\mathrm{I}}_{13/2}$). The ${\mathrm{Pr}}^{3+}$ ion has been proved to be an efficient deactivated ion with the energy level of ${\mathrm{Pr}}^{3+}:{{}^3\mathrm{F}}_4$ close to ${\mathrm{Er}}^{3+}:{{}^4\mathrm{I}}_{13/2}$ ^[13–15]. Co-doping of ${\mathrm{Er}}^{3+}$ and ${\mathrm{Pr}}^{3+}$ has been regarded as a promising alternative to quench the lower level ${{}^4\mathrm{I}}_{13/2}$. Due to the energy transfer process between ${\mathrm{Er}}^{3+}:{{}^4\mathrm{I}}_{13/2}$ and ${\mathrm{Pr}}^{3+}:{{}^3\mathrm{F}}_4$, the terminal laser level (${\mathrm{Er}}^{3+}:{{}^4\mathrm{I}}_{13/2}$) can be efficiently depopulated. Thus, the laser can operate at lower ${\mathrm{Er}}^{3+}$ ion concentrations and significantly reduce the extra heat generation caused by multiphoton relaxations. Besides, the host material also plays a great role in the MIR laser emission, which should have low phonon energy, high optical transmission at 2.7 µm, minimal ${\mathrm{H}}_2\mathrm{O}$ absorption, and large radiative emission rate. Compared with the oxide material [$846{\mathrm{cm}}^{-1}$ for ${\mathrm{Y}}_3{\mathrm{Al}}_5{\mathrm{O}}_{12}$ (YAG)], the fluorides have low phonon energy [$447{\mathrm{cm}}^{-1}$ for ${\mathrm{LiYF}}_4$ (YLF)], which can suppress the non-radiative decay from the upper level ${{}^4\mathrm{I}}_{11/2}$ to the lower level ${{}^4\mathrm{I}}_{13/2}$, thus enhancing the MIR radiative efficiency^[16,17]. Besides, negative thermal dependence of the refractive index can efficiently reduce the thermal lensing effect.

In this paper, an ${\mathrm{Er}}^{3+}$ and ${\mathrm{Pr}}^{3+}$ co-doped YLF (Er,Pr:YLF) crystal has been successfully grown by the Bridgman method. The properties of absorption and emission spectra of the Er,Pr:YLF crystal were measured and analyzed. Based on the Judd–Ofelt (J-O) theory, the emission cross section and energy transfer efficiency between ${\mathrm{Pr}}^{3+}:{{}^3\mathrm{F}}_4$ and ${\mathrm{Er}}^{3+}:{{}^4\mathrm{I}}_{13/2}$ were calculated. Furthermore, a diode-end-pumped continuous wave (CW) Er,Pr:YLF laser operating at 2659 nm was realized for the first time, to the best of our knowledge. All of these results indicate that the Er,Pr:YLF crystal has great potential for 2.7 µm laser generation.

The Er,Pr:YLF crystal was grown by the Bridgman method with the initial materials of 99.99% pure LiF, ${\mathrm{YF}}_3$, ${\mathrm{ErF}}_3$, and ${\mathrm{PrF}}_3$ with the molar ratio of 51.5:44.2:4:0.3^[18]. The concentrations of ${\mathrm{Er}}^{3+}$ and ${\mathrm{Pr}}^{3+}$ in the Er,Pr:YLF crystal were determined to be 3.98% and 0.13%, respectively. Figure 1(a) shows the absorption spectrum with the range of 300–2400 nm measured by a U-3500 Hitachi fluorescence spectrophotometer under room temperature. The absorption peak located at 969 nm (${{}^4\mathrm{I}}_{15/2}\to {{}^4\mathrm{I}}_{11/2}$ transition) with a full width at half-maximum (FWHM) of $\sim 14\mathrm{nm}$ is shown in the inset of Fig. 1. Therefore, it can be pumped by the commercial $\sim 970\mathrm{nm}$ InGaAs LD. The absorption coefficient can be calculated by the following formula: $$\alpha =\frac{D(\lambda )}{l\times \lg e},$$(1)where $D(\lambda )$ and $l$ are the absorbance and the length of the crystal, respectively. The absorption coefficient at 969 nm was calculated to be $\sim 1.86{\mathrm{cm}}^{-1}$. Considering the relationship between the absorption coefficient and absorption cross section, ${\sigma }_{\mathrm{abs}}=\alpha /N$, where $N$ means the ${\mathrm{Er}}^{3+}$ ion concentration. The absorption cross section was calculated to be $0.297\times {10}^{-20}{\mathrm{cm}}^2$ at the central wavelength of 969 nm.

Figure 1(b) shows the fluorescence spectrum of Er,Pr:YLF with the spectral range of 2400–3000 nm, which was measured by Edinburgh Instruments (FLS920 and FSP920 spectrophotometers) excited by a 968 nm laser at room temperature. Two typical emission peaks located at 2685 and 2804 nm were observed. With the room temperature absorption spectra, based on the J-O theory, three typical J-O intensity parameters were calculated to be ${\mathrm{\Omega }}_2=1.77\times {10}^{-20}{\mathrm{cm}}^2$, ${\mathrm{\Omega }}_4=3.73\times {10}^{-20}{\mathrm{cm}}^2$, and ${\mathrm{\Omega }}_6=4.67\times {10}^{-20}{\mathrm{cm}}^2$, respectively^[19–26]. The spectroscopic quality factor $X={\mathrm{\Omega }}_4/{\mathrm{\Omega }}_6$ was determined to be 0.80, which indicated that the major laser transitions were significantly strong. For the ${{}^4\mathrm{I}}_{11/2}\to {{}^4\mathrm{I}}_{13/2}$ transition, the radiative transition probability ($A_{J{J}^{\prime }}$), radiative lifetime of the ${{}^4\mathrm{I}}_{13/2}$ level (${\tau }_{\mathrm{rad}}$), and fluorescence branching ratio (${\beta }_{J{J}^{\prime }}$) can be obtained to be $43.145{\mathrm{s}}^{-1}$, 3.70 ms, and 0.121, respectively. The ${{}^4\mathrm{I}}_{13/2}$ level lifetime was much shorter than that of the Er:YLF crystal (14.28 ms), indicating the effective deactivated function of ${\mathrm{Pr}}^{3+}$ ions.

The emission cross section (${\sigma }_{\text{em}}$) of the crystal can be calculated by the Fuchtbauer–Ladenburg (F-L) equation: $${\sigma }_{\text{em}}=\frac{{\lambda }^5{A}_{JJ\prime }\mathit{I}(\lambda )}{8\pi c{n}^2\int \lambda \mathit{I}(\lambda )\mathrm{d}\lambda },$$(2)where $\mathit{I}(\lambda )/\int \lambda \mathit{I}(\lambda )\mathrm{d}\lambda$ is the normalized line shape function of the measured emission spectrum. The stimulated emission cross section was calculated to be $1.834\times {10}^{-20}{\mathrm{cm}}^2$ at 2685 nm and $1.359\times {10}^{-20}{\mathrm{cm}}^2$ at 2804.6 nm, which were higher than that of Er:YLF ($1.2\times {10}^{-20}{\mathrm{cm}}^2$), $\mathrm{Er}:{\mathrm{Y}}_3{\mathrm{Sc}}_2{\mathrm{Ga}}_3{\mathrm{O}}_{12}$ (Er:YSGG, $0.43\times {10}^{-20}{\mathrm{cm}}^2$), and Er:YAG ($0.45\times {10}^{-20}{\mathrm{cm}}^2$)^[26–28]. The results show that the co-doped ${\mathrm{Pr}}^{3+}$ ions effectively enhance the 2.7 µm laser emission and have enormous potentiality in realizing low-threshold and high-power laser operation.

The energy transfer efficiency (${\eta }_{\tau }$) of ${\mathrm{Er}}^{3+}:{{}^4\mathrm{I}}_{13/2}\to {\mathrm{Pr}}^{3+}:{{}^3\mathrm{F}}_4$ is an important factor to assess the effects of co-doped ${\mathrm{Pr}}^{3+}$ and can be calculated by the following equation: $${\eta }_{\tau }=1-\frac{{\tau }_{\mathrm{rad}}}{{\tau }_{\text{Er}}},$$(3)where ${\tau }_{\text{Er}}$ is the theoretical lifetime of ${{}^4\mathrm{I}}_{13/2}$ in Er:YLF, which was calculated to be 14.28 ms^[22]. Therefore, the energy transfer efficiency was calculated to be 74.1%, indicating that co-doped deactivator ${\mathrm{Pr}}^{3+}$ ions can effectively reduce the ${{}^4\mathrm{I}}_{13/2}$ lifetimes of ${\mathrm{Er}}^{3+}$ by resonant energy transferring.

Considering the beneficial spectral characteristics, a CW laser operation was realized. The experimental setup is shown in Fig. 2. A compact concave-plane cavity was designed with the cavity length of 14 mm. A fiber-coupled 976 nm LD with a core diameter of 200 µm and a numerical aperture of 0.22 was used as the pump source. The pump light was focused onto the crystal by a focus system with a focal length of 46.5 mm and a polarization ratio of 1:1. An uncoated $a$-cut Er,Pr:YLF with dimensions of $3\mathrm{mm}\times 3\mathrm{mm}\times 10\mathrm{mm}$ was wrapped with indium foil and mounted in copper block cooled by water at a temperature of 16°C. The input concave mirror with a radius of 200 mm was high-reflection (HR) coated at 2600–3050 nm and high-transmission (HT) coated at 900–1000 nm. The plane output couplers (OCs) with two different transmissions of 1% and 3% at 2600–3050 nm were used.

Figure 3(a) shows the laser output power as a function of the absorbed pump power with different transmissions of OCs. The maximum output power of 258 mW was obtained with a slope efficiency of 7.4%. The laser threshold was as low as 52 mW with an OC of 1%. However, the maximum laser output power 258 mW is lower than the reported value (1.1 W) for Er:YLF, which may be caused by the fact that the shortened lifetime of the upper level ${{}^4\mathrm{I}}_{11/2}$ is detrimental to energy storage during diode CW pumping^[29]. Thus, the concentration ratio of ${\mathrm{Er}}^{3+}$ to ${\mathrm{Pr}}^{3+}$ needs to be optimized to increase the laser output power. Besides, by improving the cooling system, the compensation on the thermal effect could further enhance the output power.

The laser output spectrum was measured using an optical spectrum analyzer containing a grating spectrometer (Omni-$\lambda$ 300, Zolix, China) and an InSb infrared detector (DInSb5-De01, Zolix, China), as shown in Fig. 3(b). The emission peak was located at 2659 nm, which is consistent with the measured fluorescence spectra. The output laser beam quality was analyzed by a laser beam profiler (NanoScan by Photoh, Inc). Figure 4 shows the measured laser beam quality factor ${\mathit{M}}^2$ and beam profile at the maximum output power. The output laser was operating in the single transverse electromagnetic ${\mathrm{TEM}}_{00}$ mode with a beam quality ${\mathit{M}}^2$ factor measured to be ${\mathit{M}}_{\mathit{x}}^2=1.29$ and ${\mathit{M}}_{\mathit{y}}^2=1.25$ in the horizontal and vertical directions. The output laser was measured to be linearly polarized and parallel to the $c$ axis of the crystal with a polarization ratio of 9:1. The output laser stability was measured to be $\pm 1.2\%$ for 3 h of operation.

In conclusion, the spectroscopic and laser properties of Er,Pr:YLF crystals were studied. The absorption and fluorescence spectra were measured and analyzed by the J-O theory. The absorption cross section at 969 nm was calculated to be $0.297\times {10}^{-20}{\mathrm{cm}}^2$, while the emission cross section was determined to be $1.834\times {10}^{-20}{\mathrm{cm}}^2$ at 2685 nm and $1.359\times {10}^{-20}{\mathrm{cm}}^2$ at 2804.6 nm, respectively. Besides, the energy transfer efficiency from ${\mathrm{Er}}^{3+}:{{}^4\mathrm{I}}_{13}$ to ${\mathrm{Pr}}^{3+}:{{}^3\mathrm{F}}_4$ was calculated to be 74.1%, indicating the effective deactivated function of the ${\mathrm{Pr}}^{3+}$ ion. Moreover, a diode-end-pumped CW Er,Pr:YLF laser operating at 2659 nm was realized for the first time, to the best of our knowledge. A maximum output power of 258 mW is obtained with a slope efficiency of 7.4%. The higher-efficiency, higher-power Er,Pr:YLF CW lasers are expected by optimizing the concentration of ${\mathrm{Er}}^{3+}$ and ${\mathrm{Pr}}^{3+}$ ions. Our work demonstrates that the Er,Pr:YLF crystal should be a promising alternative for high-power and high-efficiency MIR laser generation.